8 replies to this topic

[Dan_I]

   Hi,

 

I am currently in the design phase of my next Newtonian astrograph (10" f/4.7 with sandwich carbon tube and Wynne corrector). While much has been discussed about mirror cell design, I feel that the secondary mirror holder has not received enough attention, in particular in the context of imaging. 

 

Holding the secondary mirror properly is mechanically complex since there are many degrees of freedom that should be both finely adjustable and lockable (hereafter by secondary optical center I mean the point on the aluminized surface of the elliptic secondary mirror which is offset from the geometrical center by the proper amount): 

1. motion of the secondary optical center along the optical axis of the primary
2. motion of the secondary optical center in the plane orthogonal to the optical axis of the primary (two degrees of freedom)
3. rotation of the secondary around the optical axis of the primary (with the secondary optical center fixed)
4. tilt of the secondary optical surface with the secondary optical center fixed (two degrees of freedom).

So overall there are six d.o.f. to deal with (three translations, three rotations). While 2. can be addressed with the centering of the secondary holder, using the tensioning nuts of the spider assembly, the remaining four d.o.f. should be taken care of by the secondary holder.

 

The common design found in most commercial telescopes (central pull screw, three collimation push screws) has two main issues:

• rotation of the secondary (point 3.) can be adjusted roughly, not finely, and cannot be locked. It is just held by friction of the collimation screws tips onto the secondary holder flat surface. 
• tilt motion of the secondary (point 4.) cannot be done independently since adjusting the collimation screws does NOT preserve the position of the secondary optical center.

As a result, secondary collimation is not as stable as it should, and adjustment is not straightforward (the effect of the collimation screws is a bit unpredictable).

 

There are also mechanical issues like resistance to vanes torsion and to the torque created by the one-sided secondary mirror load.

 

I am aware of alternate designs (such as pushed forward by MitchAlsup on this forum, or the Strock design) but, as far as I understand, they don't allow to adjust some of the degrees of freedom (1. and 3.) hence to precisely dial field illumination and tilt (which is crucial for imaging, not really for visual).

 

Are there other designs that address these issues (commercially available or not)?

 

Thanks,

 

Dan

Edited by Dan_I, 25 May 2023 - 05:17 AM.

[totvos]

You should search this forum for wire spider designs and tests.

[Dan_I]

You should search this forum for wire spider designs and tests.

 

Thanks Tom. As far as I understand, the goal of wire spider designs is to reduce diffraction effects, which is not my main concern here (I actually like diffraction spikes !).

 

I am fine with a standard spider (I plan to use a 4-vanes 1.5mm polished stainless steel, openwork) but  don't have a good design for the collimation/holding system.

Edited by Dan_I, 25 May 2023 - 07:15 AM.

[Oberon]

I do have a design somewhere and some parts made up for a spider whose tip tilt adjustments rotate around the optical axis, which eliminates most alignment issues. Naturally it is also designed to eliminate any sort of joint that may flop, flex or be bistable, consequence being that reaching perfection is now the enemy of the good.

 

Also…I like polished stainless steel, but 1.5mm thick seems excessive. I see no reason to be thicker than 0.2mm stainless.

Edited by Oberon, 25 May 2023 - 07:43 AM.

[Pinbout]

here's a 2ndry design

 

https://3dwarehouse....65/2ndry-holder

Edited by Pinbout, 25 May 2023 - 08:12 AM.

[MitchAlsup]

I am currently in the design phase of my next Newtonian astrograph (10" f/4.7 with sandwich carbon tube and Wynne corrector). While much has been discussed about mirror cell design, I feel that the secondary mirror holder has not received enough attention, in particular in the context of imaging. 

 

Holding the secondary mirror properly is mechanically complex since there are many degrees of freedom that should be both finely adjustable and lockable (hereafter by secondary optical center I mean the point on the aluminized surface of the elliptic secondary mirror which is offset from the geometrical center by the proper amount): You forgot stiffness. You want everything that is supposed to move, to move with minmal friction, and you want everythign that is not supposed to more be as stiff as possible.

1. motion of the secondary optical center along the optical axis of the primary
2. motion of the secondary optical center in the plane orthogonal to the optical axis of the primary (two degrees of freedom)
3. rotation of the secondary around the optical axis of the primary (with the secondary optical center fixed)
4. tilt of the secondary optical surface with the secondary optical center fixed (two degrees of freedom).

So overall there are six d.o.f. to deal with (three translations, three rotations). While 2. can be addressed with the centering of the secondary holder, using the tensioning nuts of the spider assembly, the remaining four d.o.f. should be taken care of by the secondary holder. I have found it not-difficult to use the collimation screws in my secondary holder (see below)

 

The common design found in most commercial telescopes (central pull screw, three collimation push screws) has two main issues: Aside from the fact that they are not stiff.

• rotation of the secondary (point 3.) can be adjusted roughly, not finely, and cannot be locked. It is just held by friction of the collimation screws tips onto the secondary holder flat surface. 
• tilt motion of the secondary (point 4.) cannot be done independently since adjusting the collimation screws does NOT preserve the position of the secondary optical center.
• ​The secondary optical center moves with collimation adjustments.

As a result, secondary collimation is not as stable as it should, and adjustment is not straightforward (the effect of the collimation screws is a bit unpredictable).​ Nor is it particularly smooth, light to the touch, or stiff.

 

There are also mechanical issues like resistance to vanes torsion and to the torque created by the one-sided secondary mirror load. Bad vane architecture begets vibration.

 

I am aware of alternate designs (such as pushed forward by MitchAlsup on this forum, or the Strock design) but, as far as I understand, they don't allow to adjust some of the degrees of freedom (1. and 3.) hence to precisely dial field illumination and tilt (which is crucial for imaging, not really for visual).

 

Are there other designs that address these issues (commercially available or not)?

 

Thanks,

 

Dan

If you want stiffness like a good set of binoculars, you need to go back to fundamentals.

          Vanes need >-< architecture

          Collimation needs to articulate the secondary at the optical axis

          The upper assembly should be a circular I-beam

          There is no flexure in the focuser and mounting bracket

 

My 13" F/3 and my 20" can be moved without seeing any vibration in the image (through an EP). This is because the whole scope is as stiff as a good set of binoculars.

[Dan_I]

I do have a design somewhere and some parts made up for a spider whose tip tilt adjustments rotate around the optical axis, which eliminates most alignment issues. Naturally it is also designed to eliminate any sort of joint that may flop, flex or be bistable, consequence being that reaching perfection is now the enemy of the good.

 

Also…I like polished stainless steel, but 1.5mm thick seems excessive. I see no reason to be thicker than 0.2mm stainless.

 

Thanks! I'd glad to see your design.  Regarding vane thickness: is 0.2mm thick enough to hold a 90mm secondary without any flex? 1.5mm seems to be a rather standard choice for astrographs.

 

 

          Collimation needs to articulate the secondary at the optical axis

         

Thanks. I agree with this (this is my point 4.), I'd like to incoporate a kind of spherical bearing in the design to achieve this.

 

Regarding >-<  architecture, I am wondering if it obtaining absolute parallelism between the opposite vanes (which is critical for imaging, not so for visual) is not more difficult than with the >< design ?
 

Edited by Dan_I, 26 May 2023 - 04:49 AM.

[MitchAlsup]

Regarding >-<  architecture, I am wondering if it obtaining absolute parallelism between the opposite vanes (which is critical for imaging, not so for visual) is not more difficult than with the >< design ?

What I did::

a) the upper assembly is a circular I-Beam

b) I built in the offset into the vane attachment points--that is the vanes to not intersect the upper assembly at 90º intervals, but the vanes after being installed and the secondary perfectly positioned ARE at 90º to each other.

c) I did this with a life-sized drawing (in my case in the computer) positioned the secondary mirror, then positioned the secondary support to the mirror, and then drew out 90º vanes, and determined the distance from the centerline of the focuser to the vane tension point, and drilled the holes there. I measured the pointes to better than 1/100" (10 thou)

d) during vane construction, I clamped the upper assembly on my table saw, and clamped the secondary support at the proper offset and at exactly 90º to the upper assembly. Then I made one vane with 2 screw holes (matching the holder) firmly attached it to the (unmoving) secondary holder, and used this to find the perfect spot to drill the third hole in the vane.

 

 

When assembled, the vanes can only end up normal to the optical axis.

[Pinbout]

I’ve been playing around with laying my steel on the top of the top ring on the uta