Changing to curved spider vanes

[moriniboy]

When I first got this old AE newt it had a single vane support which I found very difficult to collimate.

Thanks to the wonderful Astroboot I managed to get a 4 vane spider from a 150p and this sorted out the collimation problem but it made the diffraction spikes even more evident (I'm a die hard refractor man at heart)

In early December I read an article on SGL which mentioned curved spider vanes, these are available from the US for $80 plus $50 postage and with import duties etc you would probably be looking at £100 plus.

Decided to have a go at making my own using 0.5mm x 150mm stainless steel rules from eBay at £1.50 each and here is the result (now just need to paint black).

I'm thinking I might have overdone it with the length of the vanes but we will see next time it clears up......

[John]

Thats an interesting modification. I have a different design of curved vanes in my 12" dob. I don't get diffraction spikes although I guess the diffraction is still there but spread evenly over the field of view. You may find that shorter vanes give less diffraction but it's worth experimenting to see what suits you best.

[Uplooker]

What adjustment is there for centering the secondary?

Ian

[tjr]

That looks interesting. At the risk of sounding naive, is there a chance of this giving diffraction 'pinwheel'?

[moriniboy]

What adjustment is there for centering the secondary?

Ian

None, as long as you bend the vanes to the same diameter it self centres, I have checked this and its spot on ;-)

[Uplooker]

None, as long as you bend the vanes to the same diameter it self centres, I have checked this and its spot on ;-)

Excellent news, well done. I look forward to hearing what results you get on having first light.

Ian

[moriniboy]

First light achieved (between the clouds) and as expected no diffraction spikes whatsoever.

Close scrutiny of the moon and Procyon in Canis Minor showed that the diffraction was spread evenly and gave a refractor like quality to the view.

In all a cheap and most effective mod. I can now paint the vanes matt black and this will hopefully improve the contrast further.

[Merlin]

That's a very neat job, but is it necessary to have four vanes? Wouldn't three have been enough?

[John]

That's a very neat job, but is it necessary to have four vanes? Wouldn't three have been enough?

With curved vanes you can have 3, 2 or even one but with straight vanes 4 is better than 3 I believe.

[moriniboy]

It was all down to the centre mounting, I re-used the Skywatcher one which had 4 straight vanes.

In a perfect world I would have opted for 3 as a preference as there would have been less overall diffused diffraction.

[moriniboy]

Finished.

Just painted black and carried out the milk bottle carton washer mod to the secondary.

Not the best photo I can assure you the vanes are all identical.

[JamesF]

With curved vanes you can have 3, 2 or even one but with straight vanes 4 is better than 3 I believe.

With four equally-spaced straight vanes you get four diffraction spikes.  With three equally-spaced straight vanes you get six spikes.

I guess whilst they look like spikes that perhaps gives the wrong impression of what happens.  Any straight blade-like obstacle in the optical path will cause a spike on both sides of a bright object in the field of view, whether it extends along a radius of the OTA or the diameter.  So with the four vanes each one of an opposing pair produces two diffraction spikes in exactly the same place.  With three vanes each one produces a diffraction spike either side of a bright star, so you get six in all.

In fact I think any straight edge in the optical path will do this.  It doesn't need to be thin.  It can completely block the edge of the OTA.  I think that's why you get multiple diffraction spikes in images taken using stopped-down camera lenses.  The diffraction spikes are caused by the edges of the vanes.

Nice idea for a cheap mod though.  Well done for getting it to work.

James

James

[AndyH]

Lovely job.

Whilst as you say perhaps the length may be a wee bit too much, the over-length bits, so to speak, look to be outwith the lightpath, so not interfering.

A very well done mod.

Andy. 

[John]

That looks a really neat and well finished job now

With four equally-spaced straight vanes you get four diffraction spikes.  With three equally-spaced straight vanes you get six spikes.

I guess whilst they look like spikes that perhaps gives the wrong impression of what happens.  Any straight blade-like obstacle in the optical path will cause a spike on both sides of a bright object in the field of view, whether it extends along a radius of the OTA or the diameter.  So with the four vanes each one of an opposing pair produces two diffraction spikes in exactly the same place.  With three vanes each one produces a diffraction spike either side of a bright star, so you get six in all.

In fact I think any straight edge in the optical path will do this.  It doesn't need to be thin.  It can completely block the edge of the OTA.  I think that's why you get multiple diffraction spikes in images taken using stopped-down camera lenses.  The diffraction spikes are caused by the edges of the vanes.

Nice idea for a cheap mod though.  Well done for getting it to work.

James

James

Thanks for the further explanation James. I recall reading about how it worked some time ago but I'd forgotten where.

[Ruud]

It is generally a good idea to use curved vanes. The light gets diffracted evenly in all directions when the sum of the arcs is a multiple of 180°. That allows for many possibilities:

[Duke2k]

Is there a method of determining the arc length for a given secondary to edge distance?

[Ruud]

A coincides with the edge of the obstruction, B with the edge of the primary mirror.

[Duke2k]

Thanks Ruud that helps me a lot.

[hughgilhespie]

Hi,

Looking at the very informative diagrams that Ruud posted, it occurs to me that for mechanical simplicity it would be possible to use 3 complete rings to support the secondary. I realise this would give twice the obstruction area but it would also ensure that the secondary was accurately centred and give a very rigid mounting that should hold collimation very well.

I have a half finished 12.5 inch Newtonian and the next part to make is the secondary and spider so I am tempted to give this a go. I anticipate making the rings by cutting sections of tube so it  would partly depend on finding tubing in the correct diameter to make the rings.  My experience of trying to bend metal to shape has always been a complete disaster so it would be rings or nothing!

Regards, Hugh

[John]

These are the ones fitted to my 12" Orion Optics dobsonian:

[hughgilhespie]

Hi John,

Interesting. Are those as supplied or did you fit them yourself? I would also appreciate knowing how well that set up holds secondary collimation.

Regards, Hugh

[John]

1 hour ago, hughgilhespie said:

Hi John,

Interesting. Are those as supplied or did you fit them yourself? I would also appreciate knowing how well that set up holds secondary collimation.

Regards, Hugh

Hi Hugh,

The secondary support design was offered by Orion Optics as an option for a while but I believe they only do it to special order now. I bought my scope on the used market and it was already fitted.

The scope holds collimation very well - even after being taken in my car to another observing site it only needs minor collimation adjustment to the primary tilt.

 

 

[Ruud]

I've overlooked a condition in the diagrams above, so here's a correction.

The combined angle of the arced vanes still needs to be a multiple of 180°, but also: when you translate the arcs (you may rotate them during the translation, but only over 180°) and arrange them head to tail, they must form a half circle, a whole circle, one-and-a-half circle, two circles, etcetera.

John has his vanes form two circles. That's good. His solution is a variant of the diagram on the top-right.

In some of my original diagrams the arcs cannot be arranged as a multiple of a half circle. You'd have to rotate them over other angles than 180° before that could happen. That's not good.

This means two of the original diagrams won't work. I've crossed them out here:

The best solution might be three 60° arcs. The total length of the vanes would be shorter than four straight vanes or any of the other solutions. This means less total diffraction, and the diffraction is be spread out evenly, rather than concentrated in spikes.

A straight vane, by the way, causes two diffraction spikes, perpendicular to the orientation of the vane. That's why three straight vanes arranged as on the bottom left would cause six spikes.

Curved vanes cause diffraction perpendicular to the curve. For the light to be spread evenly in all directions, the arcs of the vanes must be really precise.

[moriniboy]

This is what I'm currently making for my Terrascan binos, looks like the sum of arcs will either be 360 or 540, we will have to see.

[Ruud]

I edited my previous post because I crossed out too many designs. I did not take into account that you can rotate a vane over 180° without changing its diffraction properties.

Here is why you can rotate a vane over 180° (flip it) and have no change in its diffraction pattern. The diffraction is in the directions of the coloured segments, and these directions are the same in both cases:

 

Here is a 180° vane. Such a vane by itself will evenly diffract the light in all directions, as will a combination of a number of such vanes, no matter how they are oriented.

 

Two diagrams of one 180° vane, and two 90° vanes. The second configuration causes the least diffraction of all.  (In all diagrams, the central obstruction has a radius of 1 unit and a primary mirror of 4 units, so that you can compare the total length of the vanes per design. The longer the total vane length, the more diffraction.)

The single 180° curved vane looks like it might provide a nice stable fit for the secondary, at the second shortest vane length.