Figuring Cassegrain secondary

[kuvik]

Hi!

I'm working on a Cassegrain secondary mirror. It is the first secondary I make. Now it has still some pits on the edge, I try to keep it spherical. The concave reference is a good sphere but with bad edge. I have tested the mirror with interference test, but I have no experience to judge the result.

Please, help me estimate the surface error (ignoring the bad edge).

[davidc135]

Interesting , challenging project.

The 2ndry is looking good so far.  Fractionally convex with respect to the reference curvature, each fringe being half a wave. (as I expect you know)

How does the 2ndry's edge look if the two are offset?

Can you say more about the Cassegrain details/stats and how aspheric the secondary should be?

David

 

Edited September 10, 2021 by davidc135

[davidc135]

I had a look at the Cassegrain calculator site and found that for an arbitrary 12'' F3/9 classical Cassegrain the secondary deviation from a sphere is small so you are probably close to the finished figure.

The edge zone and centre should be high and the .71r zone low but it's subtle. What do you think?

David

[Astrobits]

Well it all depends on which cassegrain design you are making. Dall-Kirkham needs a spherical secondary and the primary to be under parabolised, Classical needs a hyperbolic secondary and fully parabolised primary and the Richey-Creitien needs both to be hyperbolic.

IF your secondary is made with clear glass, no bubbles, no strain ( check with crossed polars) and has an optically flat back ( check against reference flat) then you can figure it with a normal Foucault test from the rear. Just remember that the measured ROC and focal length will be reduced by the refractive index of the glass. If you can't do it that way then make the reference surface correct before trying the interference test.

You say the reference has a "bad edge". Which way is the "bad edge", turned up or turned down? That will give you a clue as to the overall figure on the mirror.

Otherwise it looks O.K.

Nigel

 

Edited September 10, 2021 by Astrobits

[kuvik]

Hello David! Hello Nigel!

Thank you for your replies.

The primary mirror is a non-perforated 150/450 paraboloidal mirror. Originally I have made it for RFT purposes.

I want to learn how to make hyperboloidal secondary, so I try to construct an OTA for a f/18 Nasmyth-Cassegrain. The secondary is 30 mm diameter and ROC=195 mm. If I'm correct, the conical constant must be -1.95. Now the secondarys edge has no defect. The reference has TDE, I want to fix it before hyperbolizing. The back of the glasses are just flash polished, they are not optically flat. I must hyperbolize the reference and use the fringe test. At this stage (polishing) I try to practice interpreting the fringe test and find out, how to fine adjust the ROCs and the figure. I think it will be a difficult and long project.

The primary is around l/10. If I will achieve similar fringes with the hyperbolized glasses, will it be usable?

[davidc135]

I fed your figures into the Classical Cassegrain calculator and, guessing the back focus equivalent to be 10'', got a very similar CC. This translates into a max deviation from a sphere of nearly exactly 1.5fringes at .71r so more than I had thought. (If I'm right. I'll check).

I would think wave errors on each mirror are additive so if both were out by .1 lambda the sum could be between 0 and .2l.

I think the way forward would be to assess the figure with the interference test as well as you can and finish with a whole telescope system null test using either a dependable source of collimated light in single pass or a coated optical flat and double pass. Sometimes easier said than done but if you have a good null without astigmatism then you really know the project is a success.

David

Edited September 10, 2021 by davidc135
3 fringes corrected to 1.5fringes

[davidc135]

Wave front spherical aberration for a sphere

=Aperture(ins) x 22.55/(f#)^3 = 0.775lambda but CC is -1.95 not -1 so 0.775 x 1.95 =1.51lambda.

That's wavefront after reflection so surface deviation = (1.51)/2 lambda but x2 to get 1.5 fringes (above mistake corrected).

David

Edited September 10, 2021 by davidc135

[kuvik]

Unfortunatley I'm not well equipped, but I think star test will show any important problems.

The secondary is a bit oversized, so a possible edge defect will not be a problem, and/or moving the secondary I will be able to tune the overall correction if it is necessary.

Each fringe a half wave, as you said, David. If the deviation is 1 fringe, it means, the wavefront error will be lambda/2?

[Astrobits]

Each fringe is a whole wave difference at the wavefront but as it produced in reflection then the surface error is only 1/2 wave. In use, a 1/2 wave surface error will give a 1/1 wave error on the wavefront attributable to that optical surface only.

I seem to remember reading once, long ago, that one of the American manufacturers of SCT's would assemble the optical components for a telescope and then figure a selected surface to give an acceptable image.

Nigel

[davidc135]

A fiddly job figuring the tiny secondary. What's the plan?

I'd go for the lap on a spindle at, say, 20rpm.

Below is a snapshot of one side of a 100mm Schmidt corrector figured on a spindle. Each fringe took around 10 minutes although there are slight irregularities. Worth thinking about but may not work so easily with a 30mm disc.

Looking forward to developments.

 

Edited September 13, 2021 by davidc135

[kuvik]

Thank you, Nigel!

Theoretically it is possible to correct the defect of one optical component with figuring the another mirror with an opposite error. But I don't want to to that in this project 

Yesterday I have roughly hyperbolized the reference, and I have tried to match the surfaces.

David, your advice worked. The relatively high rpm helped to work down the glass from the outer zones.

Your project is very interesting! I think those small disks are polishing tools. I use small tools too for figuring larger mirrors, but in this case I'm experimenting with petal laps.

 

 

[davidc135]

3 hours ago, kuvik said:

Thank you, Nigel!

Theoretically it is possible to correct the defect of one optical component with figuring the another mirror with an opposite error. But I don't want to to that in this project 

Yesterday I have roughly hyperbolized the reference, and I have tried to match the surfaces.

David, your advice worked. The relatively high rpm helped to work down the glass from the outer zones.

Your project is very interesting! I think those small disks are polishing tools. I use small tools too for figuring larger mirrors, but in this case I'm experimenting with petal laps.

 

 

The little discs are the weights used to centre the Newton's rings in the interference test.

A petal lap sounds a good approach for your 2ndry but in my case the 100mm lap was unmodified and the Schmidt curve was generated simply by the stroke of the glass disc on the lap.

David

[kuvik]

23 hours ago, davidc135 said:

The little discs are the weights used to centre the Newton's rings in the interference test.

A petal lap sounds a good approach for your 2ndry but in my case the 100mm lap was unmodified and the Schmidt curve was generated simply by the stroke of the glass disc on the lap.

David

Impressive! 😮

[kuvik]

Finally I have decided to make a RC, so I have figured the test plate to a little stronger hyperboloid. After several attempts, this is the best, that I can get out from this mirror.
For the hyperbolization I have used a star lap and a full size lap.
The mirror has a turned down edge, but the calculated optical diameter is 27 mm, so I have made a cardboard mask for the test, which works as a baffle too in the test tube.
This evening I have observed the shadow of Ganymedes on Jupiter with this test tube. Star test shows undercorrection, but it is normal (the primary is currently a slightly undercorrected paraboloid).
The next step is to get an aluminium coating on the secondary, and after that I will figure the primary to the correct hyperboloidal form.

Thank you David and Nigel for your help!