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
610 replies to this topic

#51 frantolmy

frantolmy

    Lift Off

  • -----
  • Posts: 13
  • Joined: 03 Mar 2014
  • Loc: Cambridge, UK

Posted 27 July 2017 - 06:04 PM

Thanks for the info Cytan, I really value your time and input on this :)

 

I've seen that page you linked before but had forgotten about it. I've always been wary of playing with the primary mirror collimation screws (in fact I've never done so) there is various conflicting advice on the forums about them and indeed the original collimation 'manual' from Astro-Tech for the GSO RC says never to adjust them at all! 

 

Anyway I got a break in the clouds and got a few minutes to do some more testing.

* This time I used Polaris and initially centred it in the OTA/sensor.

* Some wind, seeing not brilliant.

* Collimation at this point is just with the Cheshire that I did to correct it after my first attempt with the Tri-Bahtinov mask previously reported.

 

Here are 3 successive images at the centre, nothing adjusted between images:

Tri Bahtinov Centres Polaris
 
Next is each corner of the sensor, top-left, top-right, bottom-left, bottom-right:
Tri Bahtinov Corners Polaris
 
Finally I used the traditional Bahtinov mask and rotated it, again on a centred Polaris, to test the results just rotating it. 4 successive images each with the mask rotated 90deg CCW:
Bahtinov Centre 90deg rotations

 

Initial thoughts:

* Collimation at the centre looks reasonable

** though you can see the differences in successive image without any adjustments that I mentioned previously.

** The 10-4 orientation looks slightly off

* The traditional Bahtinov again doesn't appear to show any miscollimation, and rotating it (on a centred star) doesn't show it either.

* At the corners, there does appear to be issues, and it alters in different corners, though the 10-4 orientation is still the main culprit.

 

I was battling clouds so had to call it a night, so this has just been an information gathering evening, no tweaking tonight :)

I didn't get a chance to take an airy disk image. Though thanks for the tip about APT, I've downloaded the free version and will try the collimation checker tool once I get an image.

 

I'm still pondering about the interpretation, any input greatly appreciated :)

 

Cheers,

Richard


  • Xwitz likes this

#52 cytan299

cytan299

    Apollo

  • *****
  • topic starter
  • Posts: 1,313
  • Joined: 24 Nov 2014

Posted 27 July 2017 - 10:16 PM

Hi Richard,

   Thanks for doing all the work getting the data! I really appreciate the work that you are doing by putting the Tri-Bahtinov through the wringer. 

 

   I agree that the 10-4 direction is the worst and it is always bad. Also the 2-8 direction as well, but not as bad. These 2 bad directions are also seen in the 3 successive centered TriBahtinov images and gets accentuated with an off-centered star that you've shown in the subsequent images.

 

   Yes, I agree that the spikes from an ordinary Bahtinov does show that it is independent of direction. Which I think is to be expected for a well focused centered star.

   

   I think if you take a picture with the Airy disk, dollars to donuts, it will also show off-centered rings as well. With the Tribahtinov data that you've shown so far, I'm going to go out on a limb and predict that the circular shadow will be something like this (offset in the 10-4 direction, with a displacement towards the 2-8 direction). I've exaggerated the disk offsets, of course, but I hope I'm right smile.gif

 

disk.jpg

 

Like I had explained earlier, the principle of the Tri-Bahtinov is ridiculously simple and it is hard for me to believe that it does not work. But I've been surprised before smile.gif

 

cytan

 

 

Thanks for the info Cytan, I really value your time and input on this smile.gif

 

I've seen that page you linked before but had forgotten about it. I've always been wary of playing with the primary mirror collimation screws (in fact I've never done so) there is various conflicting advice on the forums about them and indeed the original collimation 'manual' from Astro-Tech for the GSO RC says never to adjust them at all! 

 

Anyway I got a break in the clouds and got a few minutes to do some more testing.

* This time I used Polaris and initially centred it in the OTA/sensor.

* Some wind, seeing not brilliant.

* Collimation at this point is just with the Cheshire that I did to correct it after my first attempt with the Tri-Bahtinov mask previously reported.

 

Here are 3 successive images at the centre, nothing adjusted between images:

 
 
Next is each corner of the sensor, top-left, top-right, bottom-left, bottom-right:
 
 
Finally I used the traditional Bahtinov mask and rotated it, again on a centred Polaris, to test the results just rotating it. 4 successive images each with the mask rotated 90deg CCW:
 

 

Initial thoughts:

* Collimation at the centre looks reasonable

** though you can see the differences in successive image without any adjustments that I mentioned previously.

** The 10-4 orientation looks slightly off

* The traditional Bahtinov again doesn't appear to show any miscollimation, and rotating it (on a centred star) doesn't show it either.

* At the corners, there does appear to be issues, and it alters in different corners, though the 10-4 orientation is still the main culprit.

 

I was battling clouds so had to call it a night, so this has just been an information gathering evening, no tweaking tonight smile.gif

I didn't get a chance to take an airy disk image. Though thanks for the tip about APT, I've downloaded the free version and will try the collimation checker tool once I get an image.

 

I'm still pondering about the interpretation, any input greatly appreciated smile.gif

 

Cheers,

Richard

 


Edited by cytan299, 27 July 2017 - 10:44 PM.


#53 Psittacula

Psittacula

    Vostok 1

  • -----
  • Posts: 104
  • Joined: 27 May 2017
  • Loc: Tokyo

Posted 27 July 2017 - 10:36 PM

Hi, frantolmy,

Thank you very much for providing us with a valuable report.
For the center focus image,
It seems that the conventional Bahtinov Mask images and the Tri-Bahtinov Mask images show that they are in focus for two directions that happen to be roughly aligned. (1 o'clock, 7 o'clock) and (3 o'clock, 9 o'clock)
On the other hand, it happens that it seems to indicate that the image of Tri-Bahtinov Mask which image does not exist in the conventional Bartinov mask in that direction is out of focus. (4 to 5 o'clock, 10 to 11 o'clock)
Therefore, by matching the existing Bahtinov Mask in the direction where the Tri-Bahtinov Mask is out of focus, the same degree of de-focus may be detected.

 

Satoru



#54 frantolmy

frantolmy

    Lift Off

  • -----
  • Posts: 13
  • Joined: 03 Mar 2014
  • Loc: Cambridge, UK

Posted 28 July 2017 - 03:30 PM

Thanks for the insight guys.

My next two tests are clear then:

 

1) Get an image of an airy disk and check the rings

2) Get and image with the Bahtinov in the 10-4 direction and see if it shows a similar out of focus to the Tri-Bahtinov in that orientation.

 

Progress, we like progress :)


  • Psittacula likes this

#55 James Cunningham

James Cunningham

    Skylab

  • *****
  • Posts: 4,056
  • Joined: 07 Aug 2010
  • Loc: Maryland

Posted 29 July 2017 - 05:28 AM

I cannot find a shop to make the cut-outs here in Maryland.  I already purchased the plastic and made mask the right size (without cutouts).  I tried to exacts knife the cutouts on card board but it's sloppy.  Any suggestions?  Thanks.

Jim



#56 iwannabswiss

iwannabswiss

    Apollo

  • *****
  • Posts: 1,477
  • Joined: 14 Feb 2014
  • Loc: Somewhere in Space

Posted 29 July 2017 - 08:59 AM

You could always get this. smile.gif  Hey, you said any suggestions.  How big does yours need to be?  You could always see if someone could 3D print one.



#57 Poochpa

Poochpa

    Mercury-Atlas

  • *****
  • Posts: 2,546
  • Joined: 12 Feb 2005
  • Loc: Buffalo Burbs, NY

Posted 29 July 2017 - 11:46 AM

I cannot find a shop to make the cut-outs here in Maryland.  I already purchased the plastic and made mask the right size (without cutouts).  I tried to exacts knife the cutouts on card board but it's sloppy.  Any suggestions?  Thanks.

Jim

 

Try local sign shops. I had one made recently.


  • Psittacula likes this

#58 James Cunningham

James Cunningham

    Skylab

  • *****
  • Posts: 4,056
  • Joined: 07 Aug 2010
  • Loc: Maryland

Posted 30 July 2017 - 04:25 AM

I called one and he said that he could do it until I sent him an image of what I wanted.  He said that it was too complex and that he might mess up the plastic!  How much did your local sign shop charge you?  Thanks.

Jim


  • germana1 likes this

#59 Poochpa

Poochpa

    Mercury-Atlas

  • *****
  • Posts: 2,546
  • Joined: 12 Feb 2005
  • Loc: Buffalo Burbs, NY

Posted 30 July 2017 - 12:08 PM

I called one and he said that he could do it until I sent him an image of what I wanted.  He said that it was too complex and that he might mess up the plastic!  How much did your local sign shop charge you?  Thanks.

Jim

$28+tax for a 10" mask. I'd try a few more sign/graphics companies. 



#60 frantolmy

frantolmy

    Lift Off

  • -----
  • Posts: 13
  • Joined: 03 Mar 2014
  • Loc: Cambridge, UK

Posted 30 July 2017 - 06:40 PM

Here we go then, latest results :)

 

First up for Satoru, the traditional Bahtinov mask in the 10-4 orientation, to see if it did indeed reveal an offset of the central spike...

Baht 10 4 Pos test Lum 1x1 2s

I think you could definitely make a case that it does.

 

And for Cytan a couple of Airy disk images.

Airy Disk Lum 1x1 5s 601
Airy Disk Lum 1x1 5s 600
 
Link to download the original .fits if that's better for analysis.
 
I've tried to use the collimation tool in Apt, but am finding it hard to definitively tell if the central obstruction is off centre; I think it is...
 
Starting to get experimental results that align with the theory perhaps? Gotta love science in action :)

  • Psittacula likes this

#61 cytan299

cytan299

    Apollo

  • *****
  • topic starter
  • Posts: 1,313
  • Joined: 24 Nov 2014

Posted 30 July 2017 - 10:56 PM

Hi Richard,

   I've used the collimation tool to line up 2 circles centered about the obstruction. Unfortunately, the circles are in white frown.gif, but if you zoom in you can see them. The small circle is used to centre about the obstruction. The 2nd circle is expanded so that it touches the outermost circle. Clearly, the top left is more compressed than the bottom like I had predicted smile.gif (Whew!) I think the 2nd picture shows it more clearly.

 

ring1.jpg

ring2.jpg

 

I think even by eyeball, like you observed, the rings are not concentric but compressed towards top left.

 

cytan


  • frantolmy, gfstallin and Psittacula like this

#62 frantolmy

frantolmy

    Lift Off

  • -----
  • Posts: 13
  • Joined: 03 Mar 2014
  • Loc: Cambridge, UK

Posted 31 July 2017 - 01:20 AM

Cytan,

 

Yes I can see it too. I was using too many circles which was making it harder to see. I'll go with just 2 in future, thanks.

So the mask is agreeing with theory :) Progress!

It's nice to see that it isn't off by much at all after collimation with the Cheshire,

Next step then is to try and use it again to collimate, on a centred star, and see if I can get it to improve.



#63 cytan299

cytan299

    Apollo

  • *****
  • topic starter
  • Posts: 1,313
  • Joined: 24 Nov 2014

Posted 31 July 2017 - 07:31 AM

Cytan,

 

Yes I can see it too. I was using too many circles which was making it harder to see. I'll go with just 2 in future, thanks.

So the mask is agreeing with theory smile.gif Progress!

It's nice to see that it isn't off by much at all after collimation with the Cheshire,

Next step then is to try and use it again to collimate, on a centred star, and see if I can get it to improve.

Hi Richard,

   I'm looking forward to seeing your data after you collimated with the Tri-Bahtinov. Looks like you have a sweet setup to be able to take data on consecutive nights! I have to reassemble every time and I'd be lucky to have 2 nights in a row to be able to image.

 

cytan


  • frantolmy likes this

#64 Psittacula

Psittacula

    Vostok 1

  • -----
  • Posts: 104
  • Joined: 27 May 2017
  • Loc: Tokyo

Posted 31 July 2017 - 10:13 AM

Hi Richard,
Thank you very much for your careful examination. We were able to recognize again that Tri-Bahtinov Mask is a tool that can confirm that the focus is simultaneously matched in three orientations by the conventional Bahtinov Mask.

This tool may be available not only for Cassegrain and Newtonian, but also for off-axis optics such as Schiefspiegler.

Satoru


  • frantolmy likes this

#65 cytan299

cytan299

    Apollo

  • *****
  • topic starter
  • Posts: 1,313
  • Joined: 24 Nov 2014

Posted 31 July 2017 - 12:24 PM

Hi Richard,
Thank you very much for your careful examination. We were able to recognize again that Tri-Bahtinov Mask is a tool that can confirm that the focus is simultaneously matched in three orientations by the conventional Bahtinov Mask.

This tool may be available not only for Cassegrain and Newtonian, but also for off-axis optics such as Schiefspiegler.

Satoru

Hi Satoru,

   I'd be really interested to see someone try the mask out on a Newtonian. I'm sure there will need to be some sort of holder to secure it to the opening.

 

cytan



#66 Psittacula

Psittacula

    Vostok 1

  • -----
  • Posts: 104
  • Joined: 27 May 2017
  • Loc: Tokyo

Posted 02 August 2017 - 09:01 AM

Hi Cytan,

 

I own a 5 inches newtonian for beginners. I made it possible to fix the post #32 mask for C6 by hooking it with the rubber bands to the fastening screws of the spider. Although the mask pitch (for 800 mm) is slightly wider than OTA's focal length 650 mm, coarse testing may be possible.

As the weather recovers, I will test it. smile.gif

IMG_20170802_221928.jpg

IMG_20170802_222004.jpg

 

Satoru


  • frantolmy and cytan299 like this

#67 Poochpa

Poochpa

    Mercury-Atlas

  • *****
  • Posts: 2,546
  • Joined: 12 Feb 2005
  • Loc: Buffalo Burbs, NY

Posted 02 August 2017 - 11:04 AM

I tried the Tri-Bahtinov mask I had made for my 9.25" Celestron EdgeHD for the first time last night. I use this scope at its native FL (2321mm) for imaging with a cooled OSC camera. My collimation has been out and showed in recent imaging results. Last night, before using the Tri-B mask, I used a regular Bahtinov mask, and centered and focused a star. I then replaced the Bahtinov Mask with the Tri-B mask, which I lined up with the 3 adjustment screws of the secondary. The initial image with the Tri-B mask showed centered spikes at 5 and 11 o'clock, a little off-center at 3 and 9 o'clock, and severely off-center at 1 and 8 o'clock. I then spent a long-time making adjustments and periodically recentering the star.  Below are screen shots of the initial, better and best results during the adjustments.

A few questions: I found it difficult to determine which screws to adjust. With the traditional defocused-star collimation method, I could use a paint stick or my hand to see a shadow to point to the correct screw to adjust. With the Tri-B mask, I didn't see any shadow and had to use trial and error to make the adjustments. Is there a better way? Also, despite a night of good seeing, but mediocre transparency, the center spikes seemed to waiver between each exposure, making it hard to determine if they were centered.

 

Thank you for Cytan, Satoru, et. al., for this new collimation tool!

Mike

Initial:

trib uncoll.jpg

Better:

trib uncoll better.jpg

Best:

trib uncoll best.jpg

 

 


  • Psittacula likes this

#68 Psittacula

Psittacula

    Vostok 1

  • -----
  • Posts: 104
  • Joined: 27 May 2017
  • Loc: Tokyo

Posted 02 August 2017 - 11:25 AM

Hi Mike,

Try to hide most of one of the three Bahtinov mask parts that make up the Tri-bahtinov mask. Then the corresponding spikes will almost disappear. For 9 inch you may be able to achieve it by covering with a palm of your hand.


  • gfstallin likes this

#69 Poochpa

Poochpa

    Mercury-Atlas

  • *****
  • Posts: 2,546
  • Joined: 12 Feb 2005
  • Loc: Buffalo Burbs, NY

Posted 02 August 2017 - 11:56 AM

Hi Mike,

Try to hide most of one of the three Bahtinov mask parts that make up the Tri-bahtinov mask. Then the corresponding spikes will almost disappear. For 9 inch you may be able to achieve it by covering with a palm of your hand.

Thanks, I'll try that.



#70 cytan299

cytan299

    Apollo

  • *****
  • topic starter
  • Posts: 1,313
  • Joined: 24 Nov 2014

Posted 02 August 2017 - 12:54 PM

Hi Poochpa,

    Thanks for trying out the Tri-Bahtinov.

 

  1. Like Satoru suggested, covering one set of the slots should tell you which screw to turn.
  2. Also the wavering of the spikes is endemic of poor atmospherics. If you just use your Bahtinov mask and the Bahtinov mask tool in APT, you'll see that the spikes come in and out of symmetry as well. Perhaps longer exposure might help. Like 1.5 to 2 seconds. 

There's one more thing that would improve the Tri-Bahtinov, that is software that allows you to measure the intersections between all the spikes like the one that is in APT for the Bahtinov. That would require some work though ...

 

However, one last thing, if anyone is reading this, that the Tri-Bahtinov (or even the Bahtinov) might be useful to measure how good seeing conditions are by measuring the intersection of the spikes. This could be the basis for a cheap version of the sbig seeing monitor: https://www.sbscient...eing-monitor-2/

 

 

cytan

 

 

 

I tried the Tri-Bahtinov mask I had made for my 9.25" Celestron EdgeHD for the first time last night. I use this scope at its native FL (2321mm) for imaging with a cooled OSC camera. My collimation has been out and showed in recent imaging results. Last night, before using the Tri-B mask, I used a regular Bahtinov mask, and centered and focused a star. I then replaced the Bahtinov Mask with the Tri-B mask, which I lined up with the 3 adjustment screws of the secondary. The initial image with the Tri-B mask showed centered spikes at 5 and 11 o'clock, a little off-center at 3 and 9 o'clock, and severely off-center at 1 and 8 o'clock. I then spent a long-time making adjustments and periodically recentering the star.  Below are screen shots of the initial, better and best results during the adjustments.

A few questions: I found it difficult to determine which screws to adjust. With the traditional defocused-star collimation method, I could use a paint stick or my hand to see a shadow to point to the correct screw to adjust. With the Tri-B mask, I didn't see any shadow and had to use trial and error to make the adjustments. Is there a better way? Also, despite a night of good seeing, but mediocre transparency, the center spikes seemed to waiver between each exposure, making it hard to determine if they were centered.

 

Thank you for Cytan, Satoru, et. al., for this new collimation tool!

Mike

Initial:

attachicon.giftrib uncoll.jpg

Better:

attachicon.giftrib uncoll better.jpg

Best:

attachicon.giftrib uncoll best.jpg


Edited by cytan299, 02 August 2017 - 12:55 PM.

  • Psittacula likes this

#71 starbase25

starbase25

    Apollo

  • *****
  • Posts: 1,453
  • Joined: 30 Oct 2014

Posted 05 August 2017 - 04:04 PM

 

I generated a Tri-Bahtinov mask for my 9.25" EdgeHD using Satoru's mask drawing generator with the following parameters: 260mm for the outer diameter, 90mm for the inner diameter and FL of 2321mm. I then saved it as a svg file and took a look at it in Inkscape. Does it look correct? Where should I get it cut and what material and thickness to ask for? Thanks.

Mike

(File attached as pdf because I can't attach svg to a CN post)
attachicon.gifTribahtinovMask.pdf

Hi Poochpa,

   Just cut out the slots that I have colored in red.

 

attachicon.gifbahtinov.png

 

There are extra lines at the transition between the two angled slots in the file generated by Satoru because it's based on my python script. This is where my programming limits are shown. 

 

   The tribahtinov can be made by printing out the mask to form a template and then pasting it onto cardboard that has some rigidity. You then cut the slots out with the template pasted onto it. I actually laser cut mine, but the method I described works very well too.

 

cytan

 

I bet Edward Van Halen would love this !!!!  lol


  • gfstallin likes this

#72 Psittacula

Psittacula

    Vostok 1

  • -----
  • Posts: 104
  • Joined: 27 May 2017
  • Loc: Tokyo

Posted 08 September 2017 - 09:50 PM

Hi all,

 

There was another report using Tri-Bahtinov Mask at RC. However, since it is a Japanese page, I will introduce it with a link of google translate*1.

 

*1: https://translate.go...t-text=&act=url

 

In this report, the use of masks in OTAs with adjustment mechanisms other than secondary mirrors is considered. Compared to SC, RC and Newtonian generally have such a high degree of freedom of the optical axis adjustment mechanism that the user can operate, so it seems that this case occurs.

Also I finally tried Tri-Bahtinov Mask in Newtonian and experienced a similar situation.

 

Satoru


  • gfstallin likes this

#73 cytan299

cytan299

    Apollo

  • *****
  • topic starter
  • Posts: 1,313
  • Joined: 24 Nov 2014

Posted 08 September 2017 - 10:19 PM

Hi all,

 

There was another report using Tri-Bahtinov Mask at RC. However, since it is a Japanese page, I will introduce it with a link of google translate*1.

 

*1: https://translate.go...t-text=&act=url

 

In this report, the use of masks in OTAs with adjustment mechanisms other than secondary mirrors is considered. Compared to SC, RC and Newtonian generally have such a high degree of freedom of the optical axis adjustment mechanism that the user can operate, so it seems that this case occurs.

Also I finally tried Tri-Bahtinov Mask in Newtonian and experienced a similar situation.

 

Satoru

Hi Satoru

   Looks a like a successful use of the Tri-Bahtinov mask if the translator works properly. Am I right?

 

cytan



#74 Psittacula

Psittacula

    Vostok 1

  • -----
  • Posts: 104
  • Joined: 27 May 2017
  • Loc: Tokyo

Posted 09 September 2017 - 04:09 AM

Hi cytan,

Yes, the author reports that collimation by Tri-Bahtinov Mask succeeded, which is quite useful. In particular, he is reporting that it was able to detect errors he could not detect with collimating eyepieces.

He also says that it should be used in conjunction with pre-collimation by collimating eyepiece such as Cheshire type.

 

Satoru



#75 kingjamez

kingjamez

    Vanguard

  • *****
  • Posts: 2,334
  • Joined: 03 Oct 2006
  • Loc: Fairfax, VA

Posted 01 December 2017 - 09:26 PM

I made a version of the tri-bahtinov mask to help me dial in the collimation on my Mak-Newt. I went with three Bahtinov masks but the straight and angled peices on opposite sides of the mask. Same principle just easier for me to draw up in CAD. I tuned the image using Maskulator and it worked wonderfully. It's really remarkable how perfectly Maskulator's simulations come out. What it predicted was exactly, I mean exactly what I saw in the eyepiece. 

 

I previously have been using software to dial in my final collimation and that takes a long time. The tri-Baht is much much faster and really quite an amazing tool when you really want to nail your collimation. Two huge thumbs up!

 

-Jim

Attached Thumbnails

  • 26F32803-20AD-48E1-8F3C-8E94C965292B.jpeg

Edited by kingjamez, 01 December 2017 - 09:26 PM.

  • Toups, cengell, brave_ulysses and 3 others like 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