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

e-180 precise collimation

  • Please log in to reply
23 replies to this topic

#1 andysea

andysea

    Fly Me to the Moon

  • *****
  • topic starter
  • Posts: 5,051
  • Joined: 03 Sep 2010
  • Loc: Seattle, WA

Posted 23 July 2021 - 02:38 AM

I was having some issues with image flatness on the e-180 and I re-read the manual. There is one paragraph that talks about precise collimation, after the standard collimation procedure. In essence it says that one should rotate the focuser with the collimating tool in it and observe how the center of the crosshairs describes a circle - I guess Takahashi knows that their focusers aren't perfect. The next step they say is to center the secondary on the imaginary center of the circle described by the crosshairs. This is actually not too difficult, in essence one has to tweak the secondary until the crosshairs move evenly within the black dot that is the reflection of the collimating eyepiece. 

After that I carefully centered the primary.

 

This seems to have helped quite a bit. I am at a point of having almost even corner stars with the ASI 6200. 

 

For those struggling with collimation it may be worth trying the same procedure as described in the manual. This is a screenshot of the aberration inspector from NINA.

I have a very small tweak of the secondary to further correct the left corners but that should be trivial.

Attached Thumbnails

  • Screen Shot 2021-07-23 at 12.15.40 AM.jpg

  • tboss70, psandelle and dron2015 like this

#2 Bean614

Bean614

    Vanguard

  • *****
  • Posts: 2,096
  • Joined: 05 Nov 2015
  • Loc: Mass.

Posted 23 July 2021 - 03:55 AM

Glad it worked out!

 

Oh, by the way, as to your mentioning  "....and I re-read the manual", WOW!!!!  What a concept!



#3 andysea

andysea

    Fly Me to the Moon

  • *****
  • topic starter
  • Posts: 5,051
  • Joined: 03 Sep 2010
  • Loc: Seattle, WA

Posted 23 July 2021 - 09:14 AM

Haha yes I know right? I ve been reading many threads regarding the collimation of this scope and nobody has mentioned that specific page in the manual. I think that’s what made the biggest difference, at least in my case.

#4 sharkmelley

sharkmelley

    Aurora

  • *****
  • Posts: 4,828
  • Joined: 19 Feb 2013
  • Loc: UK

Posted 24 July 2021 - 01:07 AM

The stars in the left column of your image still look a bit misshapen to me, compared to the stars in the right column.  I think the collimation and/or tilt can be improved further.

 

Mark



#5 andysea

andysea

    Fly Me to the Moon

  • *****
  • topic starter
  • Posts: 5,051
  • Joined: 03 Sep 2010
  • Loc: Seattle, WA

Posted 24 July 2021 - 01:07 AM

that's what I said in my original post. I am trying to correct for that right now.


  • sharkmelley likes this

#6 andysea

andysea

    Fly Me to the Moon

  • *****
  • topic starter
  • Posts: 5,051
  • Joined: 03 Sep 2010
  • Loc: Seattle, WA

Posted 24 July 2021 - 01:41 AM

Still have one wonky corner but I am not sure how to correct for just one corner of the image. It is possible that my spacing may not be quite perfect. I think this is usable.

Attached Thumbnails

  • Screen Shot 2021-07-23 at 11.37.17 PM.jpg


#7 psandelle

psandelle

    Soyuz

  • *****
  • Posts: 3,839
  • Joined: 18 Jun 2008
  • Loc: Los Angeles

Posted 24 July 2021 - 11:45 AM

Shims, Gerd Neumann CTU, etc. for tilt, probably. But you're very close for most things (which isn't easy with that scope sometimes).

 

Paul



#8 andysea

andysea

    Fly Me to the Moon

  • *****
  • topic starter
  • Posts: 5,051
  • Joined: 03 Sep 2010
  • Loc: Seattle, WA

Posted 24 July 2021 - 12:18 PM

My camera actually has a tilt plate but I measured the sensor spacing with my mitutoyo calipers and it’s square within 20~30 microns. I took multiple measurements at each corner and averaged them out. There is a slight amount of tilt in the focuser which is, I believe, what the manual addresses. Unfortunately my adapter is 2mm thick and there isn’t enough back focus for an additional tilt adapter anyway.
  • psandelle likes this

#9 andysea

andysea

    Fly Me to the Moon

  • *****
  • topic starter
  • Posts: 5,051
  • Joined: 03 Sep 2010
  • Loc: Seattle, WA

Posted 24 July 2021 - 12:40 PM

I think it’s nearly impossible to get perfectly symmetrical stars with the e180 and a sensor like the imx455. Adam Block uses the same scope but he paired it with the kaf9000. The 12 micron pixels of that sensor should be a lot more forgiving. I’d like to get one of those cameras but they are pretty rare nowadays. I think I’m pretty close now and what made the difference for me was thoroughly reading the manual and understanding the methodology. All the posts that I see that discuss the collimation of the e180 never touch on the critical point described in the last page. That’s why I wanted to bring it to everyone’s attention. Now let’s hope that the scope holds collimation for a very very long time!
  • psandelle likes this

#10 Vic Menard

Vic Menard

    Cosmos

  • *****
  • Posts: 8,332
  • Joined: 21 Jul 2004
  • Loc: Bradenton, FL

Posted 24 July 2021 - 12:54 PM

...There is one paragraph that talks about precise collimation, after the standard collimation procedure. In essence it says that one should rotate the focuser with the collimating tool in it and observe how the center of the crosshairs describes a circle - I guess Takahashi knows that their focusers aren't perfect. The next step they say is to center the secondary on the imaginary center of the circle described by the crosshairs. This is actually not too difficult, in essence one has to tweak the secondary until the crosshairs move evenly within the black dot that is the reflection of the collimating eyepiece.

That's pretty much the same procedure for verifying a collimating tool (I'm assuming you verified your collimating tool also--because if it has a tilt error it would cause the same read error through rotation). 

 

The only problem I can see with your/Takahashi's focuser alignment procedure is that it isn't "perfect" for any alignment, although it does ensure the minimal error for every focuser position. I suppose as long as the actual error isn't too large...  



#11 andysea

andysea

    Fly Me to the Moon

  • *****
  • topic starter
  • Posts: 5,051
  • Joined: 03 Sep 2010
  • Loc: Seattle, WA

Posted 24 July 2021 - 01:01 PM

Vic I have not put the collimating tool on my lathe to verify its precision but I assumed that its runout if any would be in the microns range. The takahashi tool is a Cheshire that is threaded into the focused through a visual adapter. All connections are threaded so I assumed that as my starting point.
You are correct when you say that I didn’t collimate for a specific camera angle. However when I tried that I would always get a larger error in one of the corners, it was extremely repeatable too. Given the tolerances at f2.8 I don’t think it’s actually possible to zero in on all errors unless the mechanics of the scope were completely rebuilt.
  • psandelle likes this

#12 Vic Menard

Vic Menard

    Cosmos

  • *****
  • Posts: 8,332
  • Joined: 21 Jul 2004
  • Loc: Bradenton, FL

Posted 24 July 2021 - 01:53 PM

Andy, I'm familiar with the Takahashi tool. If the tool itself has an internal tilt error, it will cause the same error you corrected when rotating the focuser. Whether it's threaded or not makes no difference--ultimately, everything must register and be locked in place when assessing the axial alignments, especially if the alignment error tolerance is very small.

 

I'm not sure that the mechanics of the scope would require a complete rebuild, but the focuser (tube interface to image plane) would seem to be the place to start. I also wonder if a large sensor with small pixels will always show some coma in the corners? Does Takahashi specify the coma "free" field diameter for a given pixel size?



#13 andysea

andysea

    Fly Me to the Moon

  • *****
  • topic starter
  • Posts: 5,051
  • Joined: 03 Sep 2010
  • Loc: Seattle, WA

Posted 24 July 2021 - 02:09 PM

I think the only way to verify any out of squareness of the Takahashi tool would be to build a threaded adapter that would then be chucked to a lathe (my lathe has a runout in the single digit micron range) then check the surface with a feeler gauge dial test indicator. I didn't do that.

My thinking is that with threaded connections, everything registers on the mating surfaces which should help keeping things square, as opposed to pressure fit connections that can easily introduce tilt.

 

The published spot size would suggest that I will always see some error at the corners with a 3.76 micron sensor. That's why I think that a 9 or 12 micron camera would be better suited. Or I could just bin my sensor of course and then drizzle the data. Seems a bit counter intuitive but It should work. 


Edited by andysea, 24 July 2021 - 02:15 PM.

  • psandelle likes this

#14 andysea

andysea

    Fly Me to the Moon

  • *****
  • topic starter
  • Posts: 5,051
  • Joined: 03 Sep 2010
  • Loc: Seattle, WA

Posted 24 July 2021 - 02:17 PM

I also wish that the Takahashi focuser were adjustable but it's not. There is no way to correct its tilt.

It's actually very accurate tho, I rotated it and measured the flatness of it's edge with a dial test indicator and it was surprisingly even so there isn't major tilt. This applied to both the focuser body and the draw tube.


  • calypsob likes this

#15 calypsob

calypsob

    Fly Me to the Moon

  • *****
  • Posts: 6,763
  • Joined: 20 Apr 2013
  • Loc: Virginia

Posted 24 July 2021 - 02:21 PM

I also wish that the Takahashi focuser were adjustable but it's not. There is no way to correct its tilt.

It's actually very accurate tho, I rotated it and measured the flatness of it's edge with a dial test indicator and it was surprisingly even so there isn't major tilt. This applied to both the focuser body and the draw tube.

How far is the focal length of the scope? If your lathe has a pass through center hole, you could chuck it up and manually rotate it facing a wall. Make a center spot on the wall and see if the cross hairs stay centered. Thats essentially how I collimate lasers in my wood lathe using a 4 jaw chuck. 



#16 andysea

andysea

    Fly Me to the Moon

  • *****
  • topic starter
  • Posts: 5,051
  • Joined: 03 Sep 2010
  • Loc: Seattle, WA

Posted 24 July 2021 - 02:23 PM

It sure does. Good thinking. I may try that. I also have another of the same tool It may be interesting to compare the two.

I have a 4 jaw chuck on the lathe right now!



#17 andysea

andysea

    Fly Me to the Moon

  • *****
  • topic starter
  • Posts: 5,051
  • Joined: 03 Sep 2010
  • Loc: Seattle, WA

Posted 24 July 2021 - 02:32 PM

I also wonder if it's really that important that the Cheshire is perfectly true. If I center the reflection of the Cheshire on the center of the imaginary circle described by the focuser when it rotates, whether the Cheshire is true or not, makes little difference.



#18 Vic Menard

Vic Menard

    Cosmos

  • *****
  • Posts: 8,332
  • Joined: 21 Jul 2004
  • Loc: Bradenton, FL

Posted 24 July 2021 - 03:13 PM

I also wonder if it's really that important that the Cheshire is perfectly true. If I center the reflection of the Cheshire on the center of the imaginary circle described by the focuser when it rotates, whether the Cheshire is true or not, makes little difference.

It depends on what you're aligning. The "Cheshire" part of the combination tool is the bright Cheshire ring alignment (relative to the primary mirror center mark), used to align the primary mirror tilt. At f/2.8, the read error tolerance is about 0.24mm, maybe 0.5mm depending on the resolution of the coma at the sensor center. The "other part" of the combination tool is the cross hair alignment (relative to the primary mirror center mark), used to align the secondary mirror tilt (focuser axis). At 180mm, the read error tolerance is about 0.27mm, maybe a little more depending on the resolution of the defocus at the sensor corner radius.

 

Of course, assessing both of these errors requires a primary mirror center mark that's very closely aligned to the optical center of the primary mirror. Given that all of these errors have direction, even a 0.5mm primary mirror center mark error can create a "lucky alignment" scenario.


Edited by Vic Menard, 24 July 2021 - 03:15 PM.


#19 Vic Menard

Vic Menard

    Cosmos

  • *****
  • Posts: 8,332
  • Joined: 21 Jul 2004
  • Loc: Bradenton, FL

Posted 24 July 2021 - 03:44 PM

Looking online, your imx455 has a 21.65mm corner radius and 3.76 micron pixels (Adam Block's kaf9000 is 26mm, but as you noted, 12 micron pixels). Even a Paracorr 2 has a 4 micron spot size at f/3 and 20mm, so you're not giving the e-180 coma corrector much room for error.

 

"The published spot size would suggest that I will always see some error at the corners with a 3.76 micron sensor. That's why I think that a 9 or 12 micron camera would be better suited. Or I could just bin my sensor of course and then drizzle the data. Seems a bit counter intuitive but It should work."

 

I guess we're in agreement.  waytogo.gif



#20 rockstarbill

rockstarbill

    Cosmos

  • *****
  • Posts: 9,168
  • Joined: 16 Jul 2013
  • Loc: Snohomish, WA

Posted 31 July 2021 - 07:08 AM

I think it’s nearly impossible to get perfectly symmetrical stars with the e180 and a sensor like the imx455. Adam Block uses the same scope but he paired it with the kaf9000. The 12 micron pixels of that sensor should be a lot more forgiving. I’d like to get one of those cameras but they are pretty rare nowadays. I think I’m pretty close now and what made the difference for me was thoroughly reading the manual and understanding the methodology. All the posts that I see that discuss the collimation of the e180 never touch on the critical point described in the last page. That’s why I wanted to bring it to everyone’s attention. Now let’s hope that the scope holds collimation for a very very long time!


Time for that 16803!
  • andysea likes this

#21 andysea

andysea

    Fly Me to the Moon

  • *****
  • topic starter
  • Posts: 5,051
  • Joined: 03 Sep 2010
  • Loc: Seattle, WA

Posted 31 July 2021 - 12:53 PM

Time for that 16803!

I've been thinking about it. The KAF9000 would be even better but very rare nowadays.



#22 rockstarbill

rockstarbill

    Cosmos

  • *****
  • Posts: 9,168
  • Joined: 16 Jul 2013
  • Loc: Snohomish, WA

Posted 31 July 2021 - 02:43 PM

I've been thinking about it. The KAF9000 would be even better but very rare nowadays.


The kaf9000 do suffer from significant RBI problems though.

#23 andysea

andysea

    Fly Me to the Moon

  • *****
  • topic starter
  • Posts: 5,051
  • Joined: 03 Sep 2010
  • Loc: Seattle, WA

Posted 31 July 2021 - 02:54 PM

There are ways to correct that:)

#24 rockstarbill

rockstarbill

    Cosmos

  • *****
  • Posts: 9,168
  • Joined: 16 Jul 2013
  • Loc: Snohomish, WA

Posted 31 July 2021 - 03:17 PM

There are ways to correct that:)


Yeah but it eats heavily into imaging time. For really long exposures is no big deal. Lots of short ones and you'll spend more time RBI flooding than imaging.


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






Cloudy Nights LLC
Cloudy Nights Sponsor: Astronomics