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

34" F2.89 Large Fork Mount BHMO Project Build

  • Please log in to reply
291 replies to this topic

#51 jtsenghas

jtsenghas

    Fly Me to the Moon

  • *****
  • Posts: 5,234
  • Joined: 14 Sep 2014
  • Loc: The flatlands of Northwest Ohio 41.11N --Bloomdale

Posted 30 December 2016 - 12:26 AM

I hope Chris doesn't mind my continuing this discussion as a slight digression on his build,  but this is a topic that I really think warrants some further consideration for such massive mirrors in general, and may help others considering building such huge scopes with their mirror cell designs. 

I trust that as long as we keep the conversation respectful and on the topic of the behavior of such supports with respect to collimation effects, that no offense will be taken. 

 

Regarding the following statements, 

.....First the flexions on the triangles will probably never be homogeneous, as accuracy needed to achieve perfect balancing and triangle support positioning on the base plate are pretty high, and if finally they are, the edge supports to mirror contact points will probably move with these (hypothetically homogeneous) flexions....

it is my position that the deflections of the all the various triangles needn't be exactly equal in magnitude to maintain tight collimation tolerances.

 

I trust it is already understood that the forces on the glass will equalize among the supports, or in the case of unequal force designs, will be closely represented by PLOP.

 

The point that I wish to make here is that a large amount of mechanical averaging is occurring among all the cell elements, even if some of the components such as these triangles may flex as much as a millimeter when the scope is pointed at zenith. Although such a large heavy mirror as this one may require a little more focus adjustment on this cell for different altitude objects than would be required if an incredibly stiff cell were used, I wouldn't expect the views to be degraded in comparison.

 

I realize that slight variation in material properties and triangle dimensions will result in some variation in how much each triangle deflects.  The same is true for the triangle supports. The amount this variation will affect primary collimation, however,  will depend on how much the AVERAGES OF DEFLECTION of all the elements attached to each of the three rearmost pivots vary.  

 

I understand that these fast scopes have primary collimation tolerances of only a fraction of a millimeter, and tilt effects of the mirror are amplified a few times to the focal length.  The amount these components flex under load will affect the amount of scope refocus required for different postures. The amount that variation in stiffness among the components will affect collimation due to tilt, however, will depend on the differences in the average deflections of all the components in each third of the the entire cell.  This averaging effect will greatly diminish the amount of variation from component variation. 

 

I think this is important with respect to cell design for these monster scopes. I don't want others to get discouraged by overdesigning the structure, and then choosing not to build a scope due to excessive cell weight in addition to mirror weight, or material costs.

 

As long as the variation of tilt keeps the aim of the mirror within collimation tolerances, and the edge support remains close enough to the center of gravity of the mirror, I expect such a cell to yield phenomenal results. I think the tendency is to overbuild to err on the side of caution by many, and sometimes at significant cost in materials and weight.

 

As for huge scopes that may flex their cell components significantly when aimed at high altitudes, compensating edge supports that remain at the center of gravity plane may be especially important. Howie Glatter's sling on linear bearings is one neat solution that I've seen on addressing this issue.

 

The main point I'm trying to make here is that a tremendous amount of dimensional averaging occurs within these large cells that reduces overall variation in collimation. Keeping components light helps to minimize variation in forces on the mirror from any torque from components whose centers of gravity and pivots aren't exactly coincident.

 

Approximating the forces to the mirror to calculated values is what really matters- -as long as collimation is kept within tolerances. 

 

My hat is off to you,  Chris!  Keep building! 


Edited by jtsenghas, 30 December 2016 - 12:40 AM.

  • stargazer193857 likes this

#52 mark cowan

mark cowan

    Vendor (Veritas Optics)

  • *****
  • Vendors
  • Posts: 9,689
  • Joined: 03 Jun 2005
  • Loc: salem, OR

Posted 30 December 2016 - 12:43 AM

The point that I wish to make here is that a large amount of mechanical averaging is occurring among all the cell elements, even if some of the components such as these triangles may flex as much as a millimeter when the scope is pointed at zenith. Although such a large heavy mirror as this one may require a little more focus adjustment on this cell for different altitude objects than would be required if an incredibly stiff cell were used, I wouldn't expect the views to be degraded in comparison.

That's pretty much the key.



#53 ctcables

ctcables

    Viking 1

  • *****
  • topic starter
  • Posts: 570
  • Joined: 14 Nov 2006
  • Loc: Oregon

Posted 30 December 2016 - 05:41 AM

J.T,

 

I understand the theory, unfortunately the reality of these big mirrors is unforgiving. The reason why professionals make mirror cell pretty stiff is not groundless either.

What can be afforded with small mirrors is no longer possible on bigger ones (actually when R4/t2 ratio becomes large, here 15800).

First the flexions on the triangles will probably never be homogeneous, as accuracy needed to achieve perfect balancing and triangle support positioning on the base plate are pretty high, and if finally they are, the edge supports to mirror contact points will probably move with these (hypothetically homogeneous) flexions. Depending on design (we don't have all details here, but the base plate is a clue), differential flexions may occur also during collimation, which may also ruin optical quality.

 

Well, this may become an endless discussion, I'm not sure I want to go deeper into it , would be more friendly face to face with a beer in hand, but this is definitely a good topic.

 

I agree on your conclusion, carry on ctcable, looking forward to your first light,

Jean-Marc

Well it is not my first mirror cell and my 20"  shows great detail on the ring at over 1400x from osp.  I am no stranger to flexion and thank you for your concern.  I have used as many tips from Mike Lockwood mirror cell design and will do my best.  I have a better understanding of precision then you might give me credit for working with automation in the semiconductor industry for 20 years now. Might not have a fancy shop but I think things will turn out just fine.  



#54 jtsenghas

jtsenghas

    Fly Me to the Moon

  • *****
  • Posts: 5,234
  • Joined: 14 Sep 2014
  • Loc: The flatlands of Northwest Ohio 41.11N --Bloomdale

Posted 30 December 2016 - 09:40 AM

Regarding the design and placement of the edge supports: 

 

I am pleased to see from an early post on this thread that you plan to use rollers with their axes perpendicular to the mirror axis to avoid having them induce loads in the direction of the mirror axis, and have no doubt your plan is to set them at the height of the center of gravity of the mirror. 

 

I would like to offer a recommendation, if I may, on tuning in your setup once you have the mirror in the cell. The back supports will see half the load of maximum (when the scope is aimed at zenith) when it is aimed at 30 degrees altitude.  Assuming you will seldom view very close to the horizon, the midpoint of their range of loading will probably occur when the scope is aimed about 35-40 degrees high. I propose that you tweak your back supports at the base of your cell or the heights of the edge roller supports so that the rollers contact the desired locations on the edge WHEN the scope is aimed 35-40 degrees high. 

 

Doing so should minimize the amount their actual heights vary from the target height (at the plane of the center of gravity of the mirror) throughout the entire range of the scope as pressure on the back supports vary.

 

It is not always recognized that the midpoint of the loading on the back supports occurs at such a low altitude by ATMs, but this load is a function of the sine of the altitude, which is 0.5 at 30 degrees. 

 

If your cell compresses 1 mm at zenith with respect to the edge supports, for example, setting your edge supports as I describe will reduce the amount the roller heights vary from the target locations to about 0.5 mm, which is less than 1/100 of your edge thickness. Whatever deflection you do have, I think it would be wise to set these dimensions at such a low altitude to minimize variation from target locations. 


Edited by jtsenghas, 30 December 2016 - 09:43 AM.

  • ctcables likes this

#55 Mike Lockwood

Mike Lockwood

    Vendor, Lockwood Custom Optics

  • *****
  • Vendors
  • Posts: 1,872
  • Joined: 01 Oct 2007
  • Loc: Usually in my optical shop

Posted 30 December 2016 - 12:50 PM

Chris and I have spoken at length about the cell design from the very start of this project.  He should have a mirror to test in the cell quote soon.

 

I am not worried at all about a slight shift of the edge support - that's what the rollers are there to allow, after all.

 

However, with an alt-azimuth design, we know very well how to build mirror cells, but since this is to be on an equatorial mount, the open question is - how will the rotation of the primary cell, and thus the rotation of the edge supports, affect the primary mirror?

 

John Pratte and I used a similar design on the 36" f/6.3 at Warren Rupp Observatory.  When I made the new mirror, it needed a new cell, so we incorporated some of our cell developments.  It seems to work well, scroll down to the 5th image for a rear view:

  http://www.loptics.c...012/HH2012.html

 

For a top view, the unfinished cell is shown in the last image on John's mirror cell page:

  http://www.jpastrocraft.com/cells.htm

 

The advantage of using composite materials for the cell is that the CTE is less than the steel cell that we used at Warren Rupp, and therefore the rollers can be placed a little closer to the mirror without having to leave room for the cell to shrink on a cold night.  This means less lateral movement of the mirror, which is fairly critical at f/2.9, as the telescope moves around the sky.


Edited by Mike Lockwood, 30 December 2016 - 12:50 PM.

  • ctcables, jtsenghas and Tyson M like this

#56 jtsenghas

jtsenghas

    Fly Me to the Moon

  • *****
  • Posts: 5,234
  • Joined: 14 Sep 2014
  • Loc: The flatlands of Northwest Ohio 41.11N --Bloomdale

Posted 30 December 2016 - 02:26 PM

I had forgotten that this will be an equatorial setup.  My particular suggestions above are for a dob mount. My error.

 

 If the rollers are all positioned with the mirror pointing straight up, they probably won't shift rearward much towards the back of the mirror with lower scope angles. If that shift is observed to be more than a millimeter with compression of the cell it might be worth tuning them to be a little high to split the difference to minimize any effect on astigmatism, but I think you are probably right that any deviations this cell will experience will be of little consequence. I realize the rollers won't provide an axial load to the mirror, but am hoping that they do remain close to the center of gravity of the mirror.

 

 Thank you, Mr. Lockwood.



#57 jtsenghas

jtsenghas

    Fly Me to the Moon

  • *****
  • Posts: 5,234
  • Joined: 14 Sep 2014
  • Loc: The flatlands of Northwest Ohio 41.11N --Bloomdale

Posted 30 December 2016 - 02:32 PM

John Pratte and I used a similar design on the 36" f/6.3 at Warren Rupp Observatory.  When I made the new mirror, it needed a new cell, so we incorporated some of our cell developments.  It seems to work well, scroll down to the 5th image for a rear view:

  http://www.loptics.c...012/HH2012.html

 

 

Wow! Hand wheels for collimation knobs!  That's one big puppy! 



#58 polaraligned

polaraligned

    Viking 1

  • -----
  • Posts: 923
  • Joined: 26 Dec 2008

Posted 30 December 2016 - 04:08 PM

I don't see any indication in prior posts that this will be an EQ mount.  Rather, I see several pictures of Alt/Az fork mounts. 



#59 mark cowan

mark cowan

    Vendor (Veritas Optics)

  • *****
  • Vendors
  • Posts: 9,689
  • Joined: 03 Jun 2005
  • Loc: salem, OR

Posted 30 December 2016 - 04:41 PM

I guess you had to have seen the other posts about it that have shown the conceptual drawings.  :shrug:

 

Although this is a clue:

 

post-20138-0-10382800-1473107526.jpg


  • ctcables likes this

#60 polaraligned

polaraligned

    Viking 1

  • -----
  • Posts: 923
  • Joined: 26 Dec 2008

Posted 30 December 2016 - 06:47 PM

Mark, the only conceptual drawing was this:

post-20138-0-19087900-1473106484_zps1gdl

 

It does not indicate an EQ mount, though of course, it could turn into one.  It looks more like a Planewave alt/az knockoff. There are 2 pictures showing alt/az mounted scopes.  Being that this is visual only, it would be better to make it alt/az.  No need for an eq.   So whatever....


Edited by polaraligned, 30 December 2016 - 06:50 PM.


#61 mark cowan

mark cowan

    Vendor (Veritas Optics)

  • *****
  • Vendors
  • Posts: 9,689
  • Joined: 03 Jun 2005
  • Loc: salem, OR

Posted 30 December 2016 - 07:04 PM

Mark, the only conceptual drawing was this:

post-20138-0-19087900-1473106484_zps1gdl

 

It does not indicate an EQ mount, though of course, it could turn into one.  It looks more like a Planewave alt/az knockoff. There are 2 pictures showing alt/az mounted scopes.  Being that this is visual only, it would be better to make it alt/az.  No need for an eq.   So whatever....

That is an equatorial mount.    The yoke rotates (it's going to be on the polar angle) and the OTA moves in elevation.

 

I suspect I've seen something you haven't seen, though...


Edited by mark cowan, 30 December 2016 - 07:05 PM.


#62 polaraligned

polaraligned

    Viking 1

  • -----
  • Posts: 923
  • Joined: 26 Dec 2008

Posted 30 December 2016 - 07:56 PM

That is an equatorial mount.    The yoke rotates (it's going to be on the polar angle) and the OTA moves in elevation.

 

I suspect I've seen something you haven't seen, though...

 

I guess if you make the **assumption** that the mount as shown will be on the polar angle, then it becomes an EQ.  The drawing does not show that.  Just because you have forks, does not make it an EQ mount.  Being visual only, it would be far more prudent to use this fork mount in alt/az mode.  Several important reasons are ease of eyepiece access, which will be far better than an EQ, and building a bearing assembly for the fork to ride on will be far easier. 

 

But, I guess until the OP clarifies it, it is an unknown.  I would strongly recommend to him that he leave it alt/az considering he plans to do outreach with it, otherwise, short of a rotating cage, the eyepiece will end up in some screwy positions. 



#63 ctcables

ctcables

    Viking 1

  • *****
  • topic starter
  • Posts: 570
  • Joined: 14 Nov 2006
  • Loc: Oregon

Posted 30 December 2016 - 07:59 PM

Some imagination needs to be used but it is not a far that hard to turn a fork mount into a eq mount. The base bearing is capable of being tilted 45deg and polar aligned. This has worked with many telescopes. I will not start it out that but want it as a option. This is not a normal build but lots of sound thought is going into it and I am not rushing it. I have been open to solution and recived help from many great people. Something I would like some help with is the best jack screw system. I want to have it powered so easy cell adjustments from the eyepiece.

#64 ctcables

ctcables

    Viking 1

  • *****
  • topic starter
  • Posts: 570
  • Joined: 14 Nov 2006
  • Loc: Oregon

Posted 30 December 2016 - 08:03 PM

Sorry on oxycodone right now. Been I'll for days now. So if I did not make sense sorry.

#65 polaraligned

polaraligned

    Viking 1

  • -----
  • Posts: 923
  • Joined: 26 Dec 2008

Posted 30 December 2016 - 08:10 PM

OK, ctcables.  That clarifies it.  It is starting life as an alt/az, but you want the option to turn it into an eq in the future.  Thanks. 

 

On Facebook, you link a video of the Planewave 1meter alt/az scope as your inspiration for building this. 


  • ctcables likes this

#66 mark cowan

mark cowan

    Vendor (Veritas Optics)

  • *****
  • Vendors
  • Posts: 9,689
  • Joined: 03 Jun 2005
  • Loc: salem, OR

Posted 30 December 2016 - 08:22 PM

 

That is an equatorial mount.    The yoke rotates (it's going to be on the polar angle) and the OTA moves in elevation.

 

I suspect I've seen something you haven't seen, though...

 

I guess if you make the **assumption** that the mount as shown will be on the polar angle, then it becomes an EQ.  The drawing does not show that.  Just because you have forks, does not make it an EQ mount.  Being visual only, it would be far more prudent to use this fork mount in alt/az mode.  Several important reasons are ease of eyepiece access, which will be far better than an EQ, and building a bearing assembly for the fork to ride on will be far easier. 

 

But, I guess until the OP clarifies it, it is an unknown.  I would strongly recommend to him that he leave it alt/az considering he plans to do outreach with it, otherwise, short of a rotating cage, the eyepiece will end up in some screwy positions. 

 

Chris has discussed this project both here and personally, so since I knew that the implementation had a huge-**** bearing on the bottom and was intended as an equatorial, in various versions I've heard of, it just never occurred to me to see it any other way.


  • polaraligned likes this

#67 jtsenghas

jtsenghas

    Fly Me to the Moon

  • *****
  • Posts: 5,234
  • Joined: 14 Sep 2014
  • Loc: The flatlands of Northwest Ohio 41.11N --Bloomdale

Posted 30 December 2016 - 09:50 PM

Chris could settle this portion of this discussion once and for all and make it moot by setting up in Santa's yard!  ;) 



#68 ctcables

ctcables

    Viking 1

  • *****
  • topic starter
  • Posts: 570
  • Joined: 14 Nov 2006
  • Loc: Oregon

Posted 10 January 2017 - 01:51 AM

first cell test complete, looks good. link to vid https://www.facebook...57763167914101/ along with some updated photos. Just need to make the mirror side supports. Cell is 14.5lbs but will add 4 more with side supports.

Attached Thumbnails

  • bearingsup.jpg
  • bearingsup2.jpg
  • celltest1.jpg
  • cellflat.jpg
  • cellweight.jpg

  • brave_ulysses likes this

#69 mark cowan

mark cowan

    Vendor (Veritas Optics)

  • *****
  • Vendors
  • Posts: 9,689
  • Joined: 03 Jun 2005
  • Loc: salem, OR

Posted 10 January 2017 - 02:43 AM

Like the new thinner mirror, Chris.  :tongue:



#70 ctcables

ctcables

    Viking 1

  • *****
  • topic starter
  • Posts: 570
  • Joined: 14 Nov 2006
  • Loc: Oregon

Posted 10 January 2017 - 09:11 PM

Like the new thinner mirror, Chris.   :tongue:

Georginna would not be happy if I turned her table into a mirror, LOL. Nice to have for testing when she is not home. 



#71 mark cowan

mark cowan

    Vendor (Veritas Optics)

  • *****
  • Vendors
  • Posts: 9,689
  • Joined: 03 Jun 2005
  • Loc: salem, OR

Posted 10 January 2017 - 10:49 PM

It's not the same size is it?  :question:



#72 polaraligned

polaraligned

    Viking 1

  • -----
  • Posts: 923
  • Joined: 26 Dec 2008

Posted 11 January 2017 - 06:48 AM

I am kinda surprised to see a 34"/2.3" thick mirror on an 18 point cell. 

How did it do in PLOP? 



#73 Mike Lockwood

Mike Lockwood

    Vendor, Lockwood Custom Optics

  • *****
  • Vendors
  • Posts: 1,872
  • Joined: 01 Oct 2007
  • Loc: Usually in my optical shop

Posted 11 January 2017 - 01:08 PM

For years Starmaster used an 18-point (moving frame) cell for a 30" x 2.0" thick mirror.  I never saw an issue and I tested those scopes quite a bit.

 

In discussing this with Chris some time ago before work was begun, I told him that for something this large, adding another layer of triangles or lots of complication was likely not worth it for the small improvement in a computer simulation (PLOP) that isn't perfectly accurate itself because it is a simulation, after all.  The difference in P-V will never be seen at the eyepiece.

 

I know the correction of mirror is quite good and smooth, and figure of revolution is good.  The outer zones came into line nicely, and its worst feature is small bump remaining in the center that might be noticed in a star test, but which won't affect images in the least.

 

So, by far the most important design goal is to build a cell that holds collimation, and for this, a mechanically simpler 18-point cell can be better and much easier to build.  After all, we're talking about a resolving power of less than 0.1 arcseconds for a 34" mirror, and any small errors in the rear support of the cell are going to be absolutely invisible.  The edge support is the real critical element for this system, especially if it is used as an equatorial instrument in the future.


  • ctcables and polaraligned like this

#74 jtsenghas

jtsenghas

    Fly Me to the Moon

  • *****
  • Posts: 5,234
  • Joined: 14 Sep 2014
  • Loc: The flatlands of Northwest Ohio 41.11N --Bloomdale

Posted 11 January 2017 - 02:14 PM

I am kinda surprised to see a 34"/2.3" thick mirror on an 18 point cell.
How did it do in PLOP?

For 2.3" thick I suspect that 18 points may yield less than RMS errors of 4 x 10^-6 mm ( 4 nanometers) error in PLOP even for 34" diameter.

I see a grainy image of a PLOP analysis on the first page of this thread but can't read the numbers, even enlarged. Chris, can you share the figures?

Edited by jtsenghas, 11 January 2017 - 02:15 PM.


#75 ctcables

ctcables

    Viking 1

  • *****
  • topic starter
  • Posts: 570
  • Joined: 14 Nov 2006
  • Loc: Oregon

Posted 11 January 2017 - 03:26 PM

 

I am kinda surprised to see a 34"/2.3" thick mirror on an 18 point cell.
How did it do in PLOP?

For 2.3" thick I suspect that 18 points may yield less than RMS errors of 4 x 10^-6 mm ( 4 nanometers) error in PLOP even for 34" diameter.

I see a grainy image of a PLOP analysis on the first page of this thread but can't read the numbers, even enlarged. Chris, can you share the figures?

 

Range is 1.79 -05 to 1.49 -06 on the plop analysis. Only slightly better with the 54pt and the 27pt had a odd pattern with about the same readings.

Attached Thumbnails

  • 18pt cell.jpg
  • 27pt cell.jpg

  • polaraligned and jtsenghas 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






Cloudy Nights LLC
Cloudy Nights Sponsor: Astronomics