578 replies to this topic

[cytan299]

Cytan and Satoru,

If I were to have a shop laser cut a mask, what material and thickness would you recommend? Thanks.

Mike

I used 3 mm thick black acrylic.

 

cytan

[Poochpa]

 

Cytan and Satoru,

If I were to have a shop laser cut a mask, what material and thickness would you recommend? Thanks.

Mike

I used 3 mm thick black acrylic.

 

cytan

 

Thanks Cytan.

Mike

[Psittacula]

Hi,
These are the Tri-Bahtinov masks I made. I pasted the pattern paper printed by the inkjet printer on cardboard about 1 mm thick and cut it with utiliy knife. The upper one is for C11 and it has support stems. The lower one is for C6.

It is light, so it will fit almost without misalignment in the SCT's corrector plate and secondary mirror cells.

 

Satoru

Edited by Psittacula, 09 June 2017 - 08:28 PM.

[cytan299]

Hi Satoru,

 

  Beautiful work!

 

cytan

 

Hi,
These are the Tri-Bahtinov masks I made. I pasted the pattern paper printed by the inkjet printer on cardboard about 1 mm thick and cut it with utiliy knife. The upper one is for C11 and it has a support stem. The lower one is for C6.

It is light, so it will fit almost without misalignment in the SCT's corrector plate and secondary mirror cells.

 

Satoru

 

Edited by cytan299, 09 June 2017 - 06:38 PM.

[Poochpa]

My wrist hurts just thinking about all the cutting you did with a utility knife to make two masks.

[frantolmy]

I have printed one of these using my 3D printer, 210mm outer diameter and 95mm inner, 3mm thick. Printed in PLA.

I haven't had a chance to test it yet due to clouds.

 

I used the web app Satoru made to generate the mask image, then Inkscape to process the SVG and finally Blender to extrude the 2D mask into a 3D model.

 

I have some images, but can't see how to post them, perhaps not permitted as I'm a new account.

 

Just wanted to thank cytan99 and Satoru for the tools to generate the mask and the inspiration.

 

Richard

[Psittacula]

Hi,

I have made a new Tri-Bahtinov mask for my C6. Assuming to use a reducer, I set the focal length as 800 mm. I made a mask by manually cutting out about 1.5 mm thick paper using an ultrasonic cutter for modeling work. However, with this creation I was extremely tired and I realize the limitations of handmade.

By the way, there was an issue that the width of the slots and uncut spaces becomes wider than the standard Bahtinov Mask design, but in the current release both the original generate.py and the webApps version have been fixed.

 

Satoru

[Psittacula]

Hi,
I've just implemented the download button on webApps to be able to store the mask drawing data more easily.

http://svg2.mbsrv.ne...i-Bahtinov.html

 

Satoru
 

[frantolmy]

Thanks for the updates to the webApp and info on the fix.

I'll be printing a new version with a smaller inner diameter and some clips to allow it to grip the secondary mirror holder on my OTA. The secondary holder is flush with the spider vanes, so unlike an SCT the mask cannot mount around the secondary, so I'll have to improvise this method to mount the mask

[frantolmy]

I hope this works to share some images of my mask so far...

 

[Psittacula]

Hi,

Because Japan was in the period of the rainy season, I could not do much test of Tri-Bahtinov Mask using actual star, but I got to report a little.

It is a test using two masks (5.3 mm and 13 mm) with different widths of slots. Seeing was not good on the day and I could not confirm the Airy disk of the star image, but I thought collimation using this mask is possible.

When Bahtinov Factor is 150, 5.3 mm corresponds to a focal length of 800 mm, and 13 mm corresponds to a focal length of 2000 mm. The actual focal length of C6 is 1500 mm.

The diffraction light streak of 5.3 mm is stronger.
This is probably because 5.3 mm mainly contributes to the spectrum of the first order, while 13 mm contributes to the 3rd order. However, when using the live image of the CMOS camera, I felt that both masks were practical.

 

 

About the order, the following posts about the original Bartinov mask will be helpful.

https://www.cloudyni...s/#entry2309839
https://www.cloudyni...-5#entry2381128
https://www.cloudyni...18#entry2513241

 

Satoru

[cytan299]

Hi Satoru,

    Great work comparing the different masks under different conditions. You're definitely much more thorough than I am .

 

    I don't use a CMOS camera, but with my main imaging camera STF8300. I then use Nebulosity "Frame and Focus" and zoom in 400x to the diffraction pattern to check both focus and collimation. I usually use Polaris as the star to focus on. IMO, 1 second integration time is good for using the mask.

 

cytan

[Psittacula]

Hi,

I finally made Tri-Bahtinov Mask with a laser cutter as a digital fabrication technology. It is a mask for C11 with a narrow pitch assuming use of reducer. The material is acrylic board of 2 mm thickness. This is difficult for me to handmade.

[cytan299]

Hi,

I finally made Tri-Bahtinov Mask with a laser cutter as a digital fabrication technology. It is a mask for C11 with a narrow pitch assuming use of reducer. The material is acrylic board of 2 mm thickness. This is difficult for me to handmade.

 

Looks great! I hope Japan gets clear soon so you can try it out.

 

cytan

[frantolmy]

I updated my model for the tri-bahtinov mask to include some clips that would attach it to the secondary mirror holder on my RC.
After a 5 hour print it came out rather nice, though still a few blemishes to work on.
Here's a few pics, followed by the meat of this post, which is a test this evening.

 

 

So there was a break in the clouds tonight and I got a chance to test.
My scope is reasonably well collimated using a cheshire, I've tried with a Glatter (RIP) laser, but had better results just by eye with the cheshire.

This first image is with a standard Bahtinov mask, looks like nice focus to me.

Here then is the result with the Tri-Bahtinov mask.

 

Whilst it looks ok in the same orientation as the Bahtinov mask; call it 8 to 2 o'clock. It definitely looks off in the other 12-6 and 10-4 orientations to me.

So I set about correcting it, I set SGP to repeatedly capture images, and used my hand to shadow each of the 3 sections of the mask to work out which corresponded to which set of spikes.
The off I set tweaking the secondary collimation screws in the usual manner.
Tiny, tiny corrections, barely a noticeable turn of a screw was needed.

Here's what I ended up with.

And the corresponding straight bahtinov mask image to complete the comparison.

To my eye I think the spikes in the Tri-Bahtinov mask certainly line up better.
A defocused star looked nice and round with an even distribution of rings.
I've still to take off the camera and confirm with a cheshire that it still looks good from that perspective, and of course take some images and see how they look.
Clouds rolled in and stopped play

Thanks for reading.

[Psittacula]

Hi frantolmy,

 

Thank you for providing a very good report.
I felt that your report gave us new insights about the Tri-bahtinov mask in the following three points.

• It is also useful for optical systems other than Schmidt Cassegrain.
• It is not much affected by four spiders.
• It is giving comparison with Cheshire eyepiece.

[frantolmy]

Satoru and cytan,

 

I'd like to upload the 3D model I've built to http://www.thingiverse.com for others to be able to print and use if they want. I would like to link to your github and webApp generator pages with acknowledgements for your invaluable help.

 

Would you mind?

 

Thanks,

Richard

[cytan299]

Satoru and cytan,

 

I'd like to upload the 3D model I've built to http://www.thingiverse.com for others to be able to print and use if they want. I would like to link to your github and webApp generator pages with acknowledgements for your invaluable help.

 

Would you mind?

 

Thanks,

Richard

Hi Richard,

   Please link to the github and webApp generator pages. Satoru has done a lot of work to improve my initial work and I think his webapp has made it possible for most people to try out the mask. 

 

   cytan

[cytan299]

I updated my model for the tri-bahtinov mask to include some clips that would attach it to the secondary mirror holder on my RC.
After a 5 hour print it came out rather nice, though still a few blemishes to work on.
Here's a few pics, followed by the meat of this post, which is a test this evening.

 

 

 

 

 

So there was a break in the clouds tonight and I got a chance to test.
My scope is reasonably well collimated using a cheshire, I've tried with a Glatter (RIP) laser, but had better results just by eye with the cheshire.

This first image is with a standard Bahtinov mask, looks like nice focus to me.

 

Here then is the result with the Tri-Bahtinov mask.

 

 

Whilst it looks ok in the same orientation as the Bahtinov mask; call it 8 to 2 o'clock. It definitely looks off in the other 12-6 and 10-4 orientations to me.

So I set about correcting it, I set SGP to repeatedly capture images, and used my hand to shadow each of the 3 sections of the mask to work out which corresponded to which set of spikes.
The off I set tweaking the secondary collimation screws in the usual manner.
Tiny, tiny corrections, barely a noticeable turn of a screw was needed.

Here's what I ended up with.

 

And the corresponding straight bahtinov mask image to complete the comparison.

 

To my eye I think the spikes in the Tri-Bahtinov mask certainly line up better.
A defocused star looked nice and round with an even distribution of rings.
I've still to take off the camera and confirm with a cheshire that it still looks good from that perspective, and of course take some images and see how they look.
Clouds rolled in and stopped play

Thanks for reading.

Hi Frantomy,

 

   Great job tweaking your collimation with the mask. It'll be interesting to see what you see with the cheshire. I think this is the first time that the mask has been tried with an RC. If your results are good, I'll be changing the title of my github page to include RC collimation as well

 

cytan

Edited by cytan299, 21 July 2017 - 07:36 PM.

[frantolmy]

Here's the link to the mask on thingiverse

https://www.thingive...m/thing:2446512

 

Next step is to take off the camera and check collimation with the Cheshire and then of course take some images and see how they look. When the cloud cooperates of course!

[frantolmy]

So the jury is still out on the mask

I took off the camera and checked the collimation with a Cheshire, it was out, by what I would consider a reasonable margin unfortunately. From prior experience certainly by enough to causes bad star shapes at the 1200mm focal length I run my RC at.

It only took a couple of tweaks to bring collimation back in line, when I rechecked with the mask it had returned to showing slightly off in one of the axis, as illustrated in the tri-bahtinov 'before' picture from my test report post above.

 

Given it was only a couple of tweaks to bring collimation back to a visual/Cheshire standard and when I attempted collimation with the mask last week I did spend a lot of time making many many tweaks, I think I can make a case that my collimation methodology with the mask may be at fault; too aggressive perhaps.

Alternatively my print of the mask may have flaws too, the print isn't perfect.

 

Time to consider my options for the next step.

 

Not giving up yet

[cytan299]

So the jury is still out on the mask

I took off the camera and checked the collimation with a Cheshire, it was out, by what I would consider a reasonable margin unfortunately. From prior experience certainly by enough to causes bad star shapes at the 1200mm focal length I run my RC at.

It only took a couple of tweaks to bring collimation back in line, when I rechecked with the mask it had returned to showing slightly off in one of the axis, as illustrated in the tri-bahtinov 'before' picture from my test report post above.

 

Given it was only a couple of tweaks to bring collimation back to a visual/Cheshire standard and when I attempted collimation with the mask last week I did spend a lot of time making many many tweaks, I think I can make a case that my collimation methodology with the mask may be at fault; too aggressive perhaps.

Alternatively my print of the mask may have flaws too, the print isn't perfect.

 

Time to consider my options for the next step.

 

Not giving up yet

Hi frantolmy,

 

I am really surprised that it would be far off w.r.t. the Cheshire. Is it possible to check collimation after tweaking with the tribahtinov using Airy rings? I don't use a Cheshire so I cannot comment on why this should not agree with it. As along as the spikes are symmetric, I don't understand why any "aggressive" tweaking spoil the results.

 

One test you can do after focusing with your Bahtinov in the 8 o'clock-2o'clock position is to rotate by 120 deg to the 10 o'clock-4 o'clock position. Is this in focus or not? If not in focus, then IMO, the scope is not collimated regardless as to whether the Cheshire says that it is.

 

Although I don't quite believe that this is the source of the problem, I cannot rule it out for now:  The mask is not lying "flat". On a Schmidt-Cassegrain, the mask will, by definition, be flat at the opening because it is lying on the corrector plate. I assume that the RC is "open". The question then is whether the mask needs to be "flat" at the RC opening w.r.t. the plane of the primary mirror. In my imagination, this should not matter, but I'm not sure.

 

I think if you can perform the tests that I suggest above, I think we can find out whether the mask will work on an RC or not.

 

cytan

 

P.S. Just a reminder to people who want to know the idea behind the tribahtinov mask: The idea is that if you focus with an ordinary bahtinov mask in one direction and then rotate it by 90 deg (or 120 deg), you should still see nice symmetric bahtinov spikes. But if they are not, then this means that there is a collimation problem because the light paths are not the same in this particular direction compared to the previous one. If the scope is well collimated, any rotation of the bahtinov mask should not show any direction that is not in focus. The tribahtinov makes it more convenient to check this because you can see the spikes in 3 directions at once.

Edited by cytan299, 25 July 2017 - 09:58 PM.

[cray2mpx]

On an SCT the star used for collimation has to be at the center of the FOV prior to adjusting the secondary. I guess this is the same with an RC? Was the star centered when checking with the Tri-Bahtinov mask?

[frantolmy]

@cytan and @cray2mpx thanks for the input and suggestions.

 

@cytan The result was a surprise to me too! After recollimating with the Cheshire I did test the Bahtinov mask in multiple positions as you suggest and it did show as being in focus regardless of the rotation. I did also try defocussing on a bright star after collimation with the Tri-Bahtinov mask and the rings looked nice and concentric to me with the central obstruction centred, though it is hard to be precise in this, trying to measure the fuzzy rings on the computer screen The mask is definitely flat with regard to sitting on the open end of the OTA, at least the mask itself is flat and it sits squarely on the secondary mirror holder.

 

@cray2mpx I think you may have hit on something here. Yes the star should be centred for collimating an RC I believe, and recentred after each adjustment of the secondary which will cause it to shift. However, I forgot this and wasn't doing it when I collimated with the Tri-Bahtinov mask. Definitely a source of user error there and I'll remove that when I next get a chance to retest.

 

I'm no expert on collimation - as we've seen - so I think I'll do some reading up too. However, one thing I did notice was that during collimation with the Tri-Bahtinov mask; generally two of the axis were ok and one had a reasonable error that I was trying to correct, but the central spike would jump around between exposures (even with no secondary adjustments). Off-centred on one exposure, more centred the next, off again the next. My theory was that this was likely bad seeing, but I don't understand why it only affected one of the axis, not all three? I know I should ideally collimate when seeing is good, but clear nights are rare here at the moment, and good seeing even rarer, 3.5" FWHM is the average. I also did try and find a single star to collimate on, not a double or variable.

 

More testing!

[cytan299]

@cytan and @cray2mpx thanks for the input and suggestions.

 

@cytan The result was a surprise to me too! After recollimating with the Cheshire I did test the Bahtinov mask in multiple positions as you suggest and it did show as being in focus regardless of the rotation. I did also try defocussing on a bright star after collimation with the Tri-Bahtinov mask and the rings looked nice and concentric to me with the central obstruction centred, though it is hard to be precise in this, trying to measure the fuzzy rings on the computer screen The mask is definitely flat with regard to sitting on the open end of the OTA, at least the mask itself is flat and it sits squarely on the secondary mirror holder.

 

@cray2mpx I think you may have hit on something here. Yes the star should be centred for collimating an RC I believe, and recentred after each adjustment of the secondary which will cause it to shift. However, I forgot this and wasn't doing it when I collimated with the Tri-Bahtinov mask. Definitely a source of user error there and I'll remove that when I next get a chance to retest.

 

I'm no expert on collimation - as we've seen - so I think I'll do some reading up too. However, one thing I did notice was that during collimation with the Tri-Bahtinov mask; generally two of the axis were ok and one had a reasonable error that I was trying to correct, but the central spike would jump around between exposures (even with no secondary adjustments). Off-centred on one exposure, more centred the next, off again the next. My theory was that this was likely bad seeing, but I don't understand why it only affected one of the axis, not all three? I know I should ideally collimate when seeing is good, but clear nights are rare here at the moment, and good seeing even rarer, 3.5" FWHM is the average. I also did try and find a single star to collimate on, not a double or variable.

 

More testing!

Hi Frantolmy,

 

 Thanks for the reply and more info. If you have concentric rings in the Airy disk after adjustments with the Tri-Bahtinov, the RC should be collimated. The results that you have written is all very mysterious to me. However, cray2mpx might be right that star has to be in the centre. FYI, I usually use Polaris centred in my scope for collimation.

 

  Also from your post above, you said that your Bahtinov saw everything focused independent of orientation. This could be true for the central part of the telescope. The tribahtinov samples three sectors of the telescope and thus can see differences between the three sectors. I found this link that may be useful: https://deepspacepla...rccollimate.php Here the author uses a Bahtinov and moves the star to different positions of the scope to check focus and thus collimation. The Tribahtinov samples three sectors instantaneously without moving the star so that you can check focus in all three sectors at the same time. (I made a mistake thinking that a single Bahtinov can check collimation by simply rotating it without moving the star. My bad )

 

 

  Please keep us posted about your results. If you take a picture of the Airy disk, I can check the collimation with APT which has a collimation checker.

 

cytan

Edited by cytan299, 26 July 2017 - 09:11 PM.