Carbon Fiber Spider

[Robert72]

Some of you may have seen my carbon fiber tube project, and I have decided to compliment this with a carbon spider. I did not intend to build a spider initially, however having sized up the spider i got with the mirror set, i found that it is slightly too short.

Obviously I could purchase a spider, but a homebrew carbon version is in keeping with the DIY theme, and gives me a chance to try out something new.

I was hoping to run the design past you fellows for a sanity check. This is not the most complex thing in the world, however there may be something i have missed as i have never actually had to set one up before.

The vanes are 0.25mm CF, bonded with 10mm section box section. the tension bolts are aluminium, the nyloc nuts are self retaining within the outer boxes, no spanner required.

The central spindle is an M6 x 80 aluminium bolt, locked into position with nyloc nuts. This should allow for rotation and +/- 5mm height adjustment.

I have designed the tilt mechanism as a single bolt adjustment type, in order to de-couple tilt from rotation (I can't see the point of the 3 bolt adjustment types).

The diagonal backplate is compressed by a spring, and pivots on 2 ball bearings. The remaining third position has a bolt in compression to do the adjustment. Hopefully this is clear from the side view.

Anyway, here is the electronic prototype, i would really appreciate it if you could let me know if i have made any omissions etc.

[Stu]

A couple of points…

What impact do the double vanes have on diffraction, must be increased surely?

Secondly, not sure how you will fully collimate the secondary with adjustment only in one direction? Unless you can guarantee it is positioned squarely then surely you need the adjustment given by three bolts, or minimum two plus a pivot.

[JamesF]

4 minutes ago, Stu said:

What impact do the double vanes have on diffraction, must be increased surely?

Yes, I'd imagine you'd get eight diffraction spikes.  I'd guess, though I'm not certain, that there'd be a small loss of contrast over a traditional four-vane spider, too.

James

[Captain Scarlet]

Assuming the pairs of vanes are parallel, you’ll get just 4 spikes, but they will be “doubled-up” in intensity over single vanes. If they’re not quite parallel, you’ll get 4 pairs of spikes, with each pair annoyingly close to each other but distinct.

Stu’s point about the single adjustment bolt is fair too: unless you can absolutely guarantee the mirror is perfectly positioned on the holder and square, you do need the extra degree of freedom provided by another bolt. 3 is actually overkill, 2 I’d say is a minimum.

M

[JamesF]

I need to make a spider for a restoration project.  My current intention is to take a cylinder of aluminium and machine out four pieces to give flats that I can make threaded holes in to fix the inner ends of the vanes in place.  With steel vanes I think the outer ends could be done with just threaded rod, slotted along part of its length to accept the vane which could be held in place with a pin at right angles to the slot.  Not sure that would work with carbon fibre though.

James

[wimvb]

13 minutes ago, JamesF said:

Yes, I'd imagine you'd get eight diffraction spikes.  I'd guess, though I'm not certain, that there'd be a small loss of contrast over a traditional four-vane spider, too.

James

@Robert72: have a look here re double vanes

http://geoastro.co.uk/zambuto2.htm

 

[Robert72]

After the height and rotation are set, then surely tilt is the only thing that requires adjustment? Adjusting rotation and tilt simultaneously with a 2/3 adjustment mechanism seems kind of un-intuitive.

I was fairly convinced of this after reading this site (from which i stole the tilt mechanism).

http://conradhoffman.com/secondary.htm

[Robert72]

10 hours ago, wimvb said:

@Robert72: have a look here re double vanes

http://geoastro.co.uk/zambuto2.htm

 

An interesting read. Thanks wimvb.

[Stu]

27 minutes ago, Robert72 said:

After the height and rotation are set, then surely tilt is the only thing that requires adjustment? Adjusting rotation and tilt simultaneously with a 2/3 adjustment mechanism seems kind of un-intuitive.

I was fairly convinced of this after reading this site (from which i stole the tilt mechanism).

http://conradhoffman.com/secondary.htm

If you can accurately set the spider so it is orthogonal in both axes across the tube, then I guess you do only need tilt in one direction.

The two/three bolts aren’t related to rotation directly (although they do allow it when slacked off sufficiently), but do allow adjustment in all directions needed. So, in my mind it all comes down to how accurately you think you can install the spider, but I may be missing something.

[wimvb]

The common 3 screw construction allows adjustment of tilt and rotation. I would think that this method gives finer rotation control than rotating the mirror by hand. If the one screw isn't perfectly aligned with the plane of incidence and reflection, then adjusting the tilt will throw the rotation off. You will need to alternate rotation and tilt adjustment.

[Robert72]

51 minutes ago, wimvb said:

The common 3 screw construction allows adjustment of tilt and rotation. I would think that this method gives finer rotation control than rotating the mirror by hand. If the one screw isn't perfectly aligned with the plane of incidence and reflection, then adjusting the tilt will throw the rotation off. You will need to alternate rotation and tilt adjustment.

Agreed. How about a 4 screw arrangement then where the 2 planes have distinct control? It may be a better way to de-couple tilt and rotation (to an extent).

[wimvb]

Do you have a sketch of this arrangement? That would make it easier to visualise.

In the common 3+1 screw configuration, two of the three screws are used for rotation/tilt adjustment. The third screw is for moving the mirror up or down (together with the central screw which holds the mirror in place). Together, the 3+1 configuration creates a push pull system, with the 3 outer screws pushing, and the central screw pulling. (But I reckon you already knew this.)

[CraigT82]

21 minutes ago, Robert72 said:

Agreed. How about a 4 screw arrangement then where the 2 planes have distinct control? It may be a better way to de-couple tilt and rotation (to an extent).

4 screws like this...?

https://www.astrosystems.biz/sechold.htm

[Robert72]

I don't have a sketch, but imagine two plates being pushed together by a spring. And then 4 screws equidistantly spaced on one of them to push them apart. If the screws protrude a few mm, then there is angle adjustment available in both planes.

[Stu]

I would think the two screws and one pivot is as simple as it gets really. That gives you adjustment in any direction needed.

[Robert72]

20 minutes ago, CraigT82 said:

4 screws like this...?

https://www.astrosystems.biz/sechold.htm

Yes basically like that.

Funnily I had already looked at this page, but due to the odd angle that the holders are in at the top left picture, I had assumed they were 3 screw.

I feel it would be too much of a challenge for me to construct a 3 screw system due to the vane positions, as the site says, "With a four screw adjustment, each screw is centered between the spider vanes, making collimation adjustments a snap".

[Astrobits]

This would not be my choice for a spider design.

Every edge produces diffraction. If you ever do a Foucault test you can see the diffraction just adjacent to the ( single ) knife edge when it is inside or outside focus. A single vane, having two edges, produces two edge diffraction patterns plus a combined diffraction where the two edges interfere. These are all superimposed on one another and cannot be seen separately so appear as one. Now double that up with the two parallel vanes and they can also interfere to produce further diffractions. You now have TEN diffractions ( each edge interferes with all the other edges as well as producing a diffraction pattern all by itself ) all superimposed on each other so will not be seen. However they will all be taking light from the centre and spreading it into the surrounding areas potentially reducing the ultimate resolution of the scope. ( Don't forget that the Airy disc and rings are formed by the edges of the primary objective and the further apart they are ( the bigger the aperture ) the finer the resolution. Conversely, the closer they are the lower the resolution and the two vanes are quite close. )

I would expect that for imaging the loss of resolution will not be apparent but attempts to split the closest doubles for the diameter of the scope might well result in a lack of success.

Nigel

[wimvb]

@Robert72: Did a little more googling on double spider vanes and came across this link

 

https://en.rbfocus.net/product-page/rb-carbon-fiber-double-spider

[Stu]

15 minutes ago, wimvb said:

@Robert72: Did a little more googling on double spider vanes and came across this link

 

https://en.rbfocus.net/product-page/rb-carbon-fiber-double-spider

@Robert72 what is the primary purpose of the scope? It seems these double vane spiders make some sense for imaging by giving greater stability but I wouldn’t want one for visual I must say.

[Robert72]

Thanks for all your input so far. The main reason for the dual vane design, is that it is so easy to design and build, although I had no idea they were so poorly regarded within the astronomy community. Back to the drawing board!

[Swoop1]

Robert- is that a true double vane spider or are we seeing open box section with the opening towards the target?

All- would what is effectively a very thick spider give increased diffraction spikes if it is box section? 

[Robert72]

5 hours ago, Swoop1 said:

Robert- is that a true double vane spider or are we seeing open box section with the opening towards the target?

All- would what is effectively a very thick spider give increased diffraction spikes if it is box section? 

It is

0.25mm sheet - 10mm box at the ends and centre - 0.25mm sheet

[Swoop1]

OK. Thanks.

[wimvb]

I dug up my copy of Maskulator and calculated the diffraction for thin single spider vanes, thick single spider vanes and thin double spider vanes (separated the width of the single thick vanes).

1. thin single

2. thick single

3. thin double

(click on each image to get the full resolution, 1024x1024 pixels)

This is highly unscientific, but it seems to me that the double vanes have stronger diffraction spikes. This is to be expected, since each edge contributes to the diffraction. There is also colour banding which is tighter (smaller structures) for thick vanes. The double vanes show a combination of colour banding for thin and thick vanes. The banding is probably related to the distance of the edges that create the diffraction. I used a double vane spacing equal to the thickness of the thick vanes.

Edited June 25, 2021 by wimvb

[Robert72]

Maskulator seems like a handy app. The double vane spacing seems quite small though (I don't know if it makes much difference).