Schiefspiegler Project

(30 June 2022)

The Schiefspiegler telescope uses two or more mirrors tilted such that the optical path is unobstructed. It was developed in the 1950s by Anton Kutter and has had many innovations since. It has a relatively long focal length (F/#) and is more suited for lunar and planetary targets than deep sky objects. I was particularly taken by a 4.25 inch design by Oscar Knab that appeared in the October 1961 issue of Sky and Telescope magazine. The two mirrors, one convex, one concave did not scare me, but the tube (especially that cut out) seemed too hard for me to make at the time.

3D Printing to the Rescue

One of the tools I've found that make designing for 3D printing so easy is the free, open source program OpenSCAD. Most of the objects we want to design consist of boxes or cylinders which are native objects in OpenSCAD. Any of the OpenSCAD objects can be manipulated (translated, rotated), added to other objects or subtracted from other objects. Thus, for example, a plate with a hole in it is a box with a cylinder subtracted from it, etc.

The image on the left below shows the problem of the cutout in the telescope tube. The vertical tube (precisely defined by an outer cylinder of 3 inches), an inner cylinder of diameter 3 - 2*wall_thickness (not shown) and a cylinder of the diameter of the light cone from the primary to the secondary mirror that is offset and tilted to the angle in the specification sheet above. The image on the right shows the result of subtracting the second two cylinders from the first. This image is converted to an STL image ready for slicing and 3D printing. All it took was about 10 lines of code.

My Ender 5 3D printer is limited to 300 mm in vertical dimension so the tube had to be printed in sections that can be glued together to make the final tube. The part of the tube with the cutout had to be split as well as shown in the two images below. The rest of the parts are simple cylinders. Ring joints to facilitate the gluing are almost trivial to design in OpenSCAD. The next image below shows the printed sections along with light stops for insertion inside the tube and a couple of the ring joints.

The .STL files for these two parts are in the Files Directory

Another tool I like and recommend is LibreCad which is a free, open source Computer Aided Design/Drafting program. It allows me to make precision drawings, make difficult measurements and visualizations. See below for dimensions and sizes for the tubes, tube junction and light paths. With the light paths I was able to specify the light stops for the tube interior and check the clearance of the mirrors, etc.

The (short) tube holding the primary mirror will be 6 inches in diameter. It must be joined to the longer tube at a precise angle and position. To that end I created another 3D model in OpenSCAD with, again, various boxes and cylinders.

The .dxf file for LibreCAD is in the Files Directory

(3 July 2022)

This is the design for the tube coupler. The print took 14 hrs, 31 minutes and about 1/3 of a spool of filament but the result is very good.

The .STL file is in the Files Folder

(3 July 2022)

This is a mock up of the smaller tube with the junction box. All the segments will be glued and the smaller tube will be painted (white, I think). The 6 inch diameter primary tube is next.

Basic OTA

(4 July 2022)

The primary tube took 6 1/2 hours to print but came out beautiful. After I verify some measurements (primary to secondary distance in particular) I will glue it all together and paint the outside.

Mirror Cells

(18 July 2022)

The primary and secondary mirror must be held securely in the OTA in very specific orientations with distinct cell designs. To aid in developing these I 3D printed dummy primary and secondary mirrors to test fit and function.

The plan diagram above shows the basic requirements. The secondary must fit in the 3 inch tube and allow the mirror to be tilted by 6.11 degrees in a plane perpendicular to the drawing. The requirement for the primary is similar with different values. The point of reference for the tilt is the center of the face of the mirror so that point must be located on the tilt axis. To meet these requirements I decided on a sort-of gimbal arrangement.

The cell that holds the mirror is a hollow tube with a pair of trunnions attached that will allow the necessary tilt. The cell will fit into an adapter ring that sits inside the respective tube and has a pair of trunnion sockets to accept the trunnions.

Secondary Cell

Tube Adapter

Assembled

Printed (with Dummy mirror installed)

(19 July 2022)

The cell and mounting for the primary mirror follows the same design with appropriate dimension changes. The tube adapters are designed to be placed at specific points within the tubes to achieve the specified distance between the centers of the two mirrors. The primary blank was 4.56 inches in diameter (what I could get at the time) but the specified diameter is 4.25 inches so the primary cell incorporates a stop of 4.25 inches located at the center of the trunnions.

Primary Cell

Primary Tube Adapter

Assembled

Printed (with dummy mirror)

The Optics

The mirrors consist of a concave primary 4.25 inches in diameter and a convex secondary of 2.17 inches in diameter. Both mirrors have a radius of curvature of 137.8 inches. As both are the same the tool used to grind the mirror can be used for the secondary (after polishing and cutting out with a diamond hole saw).

Grinding

I obtained two glass disks of 4.5 inch diameter and kinda thick, but free. I will grind them together mirror on top to get concave surface switching off to tool on top to control depth. I will use the grinding machine at Inventors Forge (just the turntable) manipulating the glass by hand. The glass will have a very short depth ( 0.016inches). The primary will be masked off to the design value of 4.25 inches to meet the design.

After an hour and a half of grinding with #80 silicon carbide, a half hour each of #120 and # 220 silicon carbide, this is what I got. This spherometer has a radius of 1.85 inches which means the depth of 0.0124 inches indicates a ROC of 137.8 inches.

The spherometer is really convenient (if you use it regularly). The base was 3D printed and the dial indicator cost about $15 on Amazon.

28 September 2022

Polishing

The fine grinding was finished out with successive wets of 25 micron, 15 micron and 9 micron aluminum oxide. The final spherometer reading was maintained by alternating the mirror on top with the tool on top.

Polishing uses a different tool called a lap. Traditional laps are made from pitch but newer artificial materials are available now so I chose one. It's not pitch but I will call it by the traditional name - pitch lap. The material will deform at room temperature to make a very close match to the mirror surface which is necessary to get a good polish and control the shape of the surface. The lap is made from a plaster disk of the same diameter as the mirror and the lap material is melted and poured on the plaster. Facets are created on the surface of the lap by pressing a steel bar into the soft material or cutting them out with a razor blade after the material has cooled. Instead of a 'grit' to remove glass we use cerium oxide suspended in water as a polishing agent. Polishing consists of moving the mirror and lap against each other with the polishing compound between them much the same as in grinding. The goal is to remove all the remaining pits from the mirror surface to produce an optically smooth surface. See image below.

Before every polishing session, the lap is well lubricated with the polishing solution and the mirror is placed on top with some weights to "cold press" the lap to achieve a really good contact between the lap and the mirror. The lap is rotated on the mirror machine while the mirror is moved back and forth across the lap. Moving the mirror may need to be done by hand as there is a lot of friction between the lap and mirror.

After about an hour of polish, the surface of the mirror is smooth enough to allow optical testing with a Foucault tester. This allows checking the form of the surface being formed and the radius of curvature. My first look is shown below where I used a home-made Ronchi grating of 60 spaces per inch. The evenly spaced lines indicate a good spherical shape in progress and a good estimate of the radius of curvature is 142 inches. My target is 137.75 inches so I need to deepen the curve during the rest of the polishing process by leaving the mirror on top of the lap.

Page updated

Report abuse