This will be the first of a series about how to optimize telescope designs using the OSLO software program. OSLO is a free download which is available from Sinclair Optics at: http://www.sinopt.com
Please email me, Steve Fejes, with questions or comments at: email@example.com
Once again, I recommend the book Telescope Optics by Rutten and van Venrooij, published by Willman-Bell. This is a good guide for understanding the different types of telescope designs. It is very helpful to have a solid foundation before you try to design and optimize telescope designs. This is a valuable reference book which takes the time to cover a lot of material which I don't plan to cover in my how to guides. Telescope Optics covers many subjects including: basic optics, telescope designs, field correctors, focal extenders and reducers, eyepieces, deviations, misalignments, tolerances, resolution, contrast, vignetting, baffles, and internal reflections.
There are several methods of optimizing with OSLO. I will give a brief description of these. Then I will give an example to get us started optimizing with OSLO. I will go step by step through the optimization process, so you can follow along. This first example will be an 8" f/15 Maksutov Cassegrain design with a separate secondary mirror. All surfaces are spherical.
After the first example of optimization, I will give a second design which has additional information about optimizing. This will show a setup for a Maksutov Cassegrain which uses an aluminized spot on the back of the corrector lens for the secondary mirror. This setup is a little different than the design we will work on first. There will be some good information in this section which should be helpful to you even if you don't want to design a Maksutov telescope. I will show how to enter curve pickups, thickness pickups, and special aperture data. These are needed to optimize a Maksutov Cassegrain design, and can be a useful tools for optimizing other types of optical designs.
The first method will be used most of the time. This is the automatic mode, where OSLO uses GENII ray aberrations to find the best design. This starts with your lens design, and then OSLO works to find progressively better solutions. GENII uses OSLO's default values to optimize your design. This type of optimizing works well in many cases, but not always. The key is to determine what lens data should be allowed to vary, and what data should not be allowed to vary. Some data should be allowed to vary, but only within certain limits which you will define. I will try to give you an idea about what data should be allowed to vary, and why. Learning this balance comes with experience. I have learned mostly by trial and error.
The first time OSLO runs an optimization routine should result in an improved design. But, this improved design might not be the best possible. So, additional attempts should be made to find better solutions. As you work with this process you will learn to better define your design goals, and to understand how one aberration will balance another aberration. To get a better solution, you might try modifying your initial design, or adjusting which items are allowed to vary.
This automatic optimization method can work very well, but sometimes the solutions are not what you expect. There are times when your original design is good, but OSLO "optimizes" to a design which is worse. This is probably a case where your priorities are different than OSLO's. There are several ways to deal with this situation. Often I find that I need to set one of the surfaces as non variable. You may need to close OSLO and then open it again and start over, especially if you have been working with OSLO for a long session.
If you cannot get OSLO to give a good optimization with the automatic method, you may want to try a different method of optimization. Or, you may want to redefine OSLO's goals for optimization, that is change the default values for GENII optimization. You can change the default values to your own parameters. This is an important method for optical engineers. I don't know how to do this, so you will have to learn how from the help information in OSLO.
I have found that OSLO's default settings give good color correction for a standard doublet refractor lens. But, OSLO does not seem to be optimized to find good apochromatic color correction in most cases. I will explore this is a different How To Guide.
The second method is the use of slider wheels. This is not an automatic method, you may vary one of several user defined variables and see the results. This lets you take control of the process. This method can be useful at times. I will cover this method of optimization in a different How To Guide.
The third method is to enter different values by hand into the surface data window. This is not as convenient as the slider wheels. But is useful at times, especially when you only want to change one value a little bit. I try this after I have a good design optimized by changing the IMS radius, or maybe one radius on a lens.
MAKSUTOV CASSEGRAIN WITH A SEPARATE SECONDARY MIRROR
I will start with a Maksutov Cassegrain design, this will have a separate secondary mirror. This is an 8 inch f/15 design. My installation of OSLO version 6 seems to do OK at optimizing this design as long as my original design has reasonable radii of curvature and spacing between optics.
If you get an error message while optimizing, just click OK. You may need to click OK many times, this will not cause problems with OLSO. It just means that OSLO cannot perform the optimization correctly. There is probably a value which is too small or large and this prevents a realistic solution. We will need to double check the surface data to see if there is an error.
You will need to create this design. This is a starting design only, it does not give very good images. I would save this design before you begin optimizing. This will allow you to go back to the starting point for a fresh start if needed. I would save each good optimization design with a different file name. I called this original design 8inF15mak1a. After I get done optimizing it for the first time I called the improved version 8inF15mak1b.
Surface 1 (AST) and surface 2 define the corrector lens. Surface 3 is the primary mirror. Surface 4 is the secondary mirror. Surface 5 is only here so the drawing will show the light coming to a focus.
It would be good to double check the data you have entered to make sure it is the same as I have shown above. Make sure you have negative values were you should have them, and positive values where needed. The lens drawing will probably show you if a value is set wrong.
The above design has three gray boxes with a "v" in them. These are the items which are allowed to vary. You set an item as variable by left clicking on the gray box to the right of the data. The first item is the RADIUS for surface 2. This window will open when you click on the gray box. In this case I have clicked on Variable (V). The small "v" will now show up in the gray box after the value -350.
You will want to do the same thing for the radius of surface 4, and the the thickness of surface 6 (IMS). Allowing the IMS thickness to vary may result in a different focal length. So, sometimes you will not want this to be variable. But, there are times when you will want it to vary some. For this example I will set it as variable.
Now that we have these three items set as variable we need to tell OSLO
how to vary them. We will use the default settings for the automatic
optimization routine. Click on: Optimize
Generate Error Function
GENII Ray Aberration
This window will open. These are the default values for GENII. Click OK on the bottom left to accept these values.
Now we are going to look at the Optimization Conditions. Click on Optimize, the Optimization Conditions.
This window will open. I have changed the default value from 0.01 Per cent improvement for continuing full iterations: to a much smaller number, 0.000000001. The value here is not too critical, but I like to have OSLO work for a better solution in the first attempt. Enter the new value in, and OSLO will accept the value as 1.0e-08, or similar based on how many zeros you have added. Click the green check in the upper left to accept this new value. You may need to click this check mark twice to accept the value and then close the window.
Now we are ready to optimize this design. Click on Optimize, then on Iterate.
You will see a window open which gives you some choices about how to iterate - optimize. You can just click on OK and accept the default values.
Now we see the results of OSLO's optimization. The values have changed in the boxes which are variable. And the optimization steps (iterations) are summarized in the text window below.
Now we will focus the new design. I have highlighted the minimum on-axis spot size (monochromatic). Click this and you will see a slight change in the IMS thickness value.
Below is the new data for our 8" f/15 Maksutov Cassegrain design. The focal length (Efl) has not changed much. It is listed in the upper right side of the Surface Data window, in the white part of the window below Notes.
EVALUATION AND FURTHER OPTIMIZATION
It is time to evaluate this design and see if it is better than the stating design. Actually, this new design gives good images and has reduced the aberrations a lot. The spot sizes look pretty good compared with the Airy disk. This design is not as good as it could be, but it is a good solution. I would now save this file with a new name.
The off axis images look better if you change the IMS to about -600mm. I found this value by entering different values in the IMS radius box. You will notice that the off axis spot sizes are now reduced. Normally it is not a good thing to allow the IMS radius to be a variable because OSLO might set the radius too short. This could interfere with the optimization, and give strange solutions. I have found it's better to modify this after the optimization is done.
You may also notice that with the IMS radius set at -600 the astigmatism graph has changed. The two lines on this graph are now about vertical.
Now that we have optimized this design once, how do we find a better solution? There are many things that could be tried to improve the design. I will list some of the possibilities.
I chose the radii for the corrector lens based on some other designs I have seen. But these starting radii may not be the optimum values. The first thing I would try is to set the radius of curvature for surface 1 at a different value. Then optimize the design again. Your evaluation of this new design will show if there is an improvement or if it is worse. I would try several such attempts, a couple with the radius shorter, then a couple with the radius longer. I would probably try changing R1 to -300mm then optimizing, and then again -275mm. I would save each new design with a new file name. Then I would try changing R1 to -350mm then optimizing, and then I would try R1 as -380mm.
Then I would evaluate these designs and see which value of R1 gives the best images, and the lowest aberrations. I would then try to narrow down the range, say the best images came with designs where R1 was -335 and -300. I would set R1 to -320 and see if this improved the images.
The spacing between the two mirrors will affect the images, especially the field curvature. The IMS radius shows the field curvature. The mirrors spacing and curves will affect the size of the secondary mirror also. You may want to examine different spacings between mirrors and adjust the back focal length for your design. I have just set a starting point for the purpose of this lesson on optimization.
In general, the best location for the corrector lens is at the center of curvature because this will reduce aberrations like coma and astigmatism. But it is more convenient to locate the corrector lens at half this distance so it can support the secondary mirror and baffle.
One thing I see is that the thickness of the corrector lens is 20mm. We should be able to reduce the aberrations if we used a thicker lens. Of course this will cost more money and take longer to reach thermal equilibrium. You might want to enter a slightly different value for the thickness, say 20.2mm or 19.8mm. Then refocus the design and evaluate it. You will see that even a change of 0.1mm will affect the images and aberrations. Modifying the thickness a little may improve the overall image quality, so it is worth investigating.
In a later How To Guide I will cover more about how to find the allowable tolerance for optical surfaces. As you can see a slight change in the corrector thickness affects the images. So does a slight change in either radius of curvature on the corrector lens.
All these surfaces are spherical, it may be possible to improve the images if we allow one surface to be slightly aspherical, say R1 which would be the easiest to test because it is concave. This should not be needed for our 8" f/15 design, but it could make a big difference with a fast focal ratio design.
SETUP FOR A DIFFERENT MAKSUTOV CASSEGRAIN
Here is a setup for a Maksutov Cassegrain which uses an aluminized spot on the back of the corrector lens for the secondary mirror. The setup is different than the design above because you will need to keep the radius of curvature and the location of the secondary mirror coincident with the back of the corrector lens. I will show you how to do this.
Enter the following data to make this lens file, then give it a name and save it. This design is based on one found in the book Telescope Optics. I have modified the thickness of the corrector a little, the original thickness is 17.347mm.
I am going to enter Special Aperture Data for surface 1. This will block the light which hits the secondary mirror obstruction. This is a refinement which will give a more accurate analysis. Click on the gray box under APERTURE RADIUS for surface 1 (AST). Click on Special Aperture Data.
This is the new window which will open. I have entered 30mm as the semi-diameter for the obstruction. The type is Ellipse, with -30 as the x min, and 30 as the x max. Then, the Action is Obstruct (gray box), with -30 Y min, and 30 Y max. Click the green check mark twice to accept these changes.
Again I will enter similar information for APERTURE RADIUS on surface 3, which is the primary mirror. This will draw the mirror with a central hole.
Enter the same -30 and 30 values as I showed on the above example.
I have not shown the Special Aperture Data window again here.
Now I will enter a Curvature pickup. This tells OSLO to keep a second surface the same as a preceding surface. In this case I will have radius 4 pickup the radius of surface 2. Surface 4 is the secondary mirror surface, while surface 2 is the back of the corrector lens. In this type of Maksutov Cassegrain these two surfaces are the same, so we need to set up the surface data like this.
Click the gray box under RADIUS for surface 4. Click on Curvature pickup.
Enter 2 in the new window. This is the pickup source surface.
click OK because we want to accept the default value of 0. The pickup constant of 0 will keep radius 4 the same as radius 2. This is what we want.
It is possible to have surface 4 be a multiple of surface 2. If we wanted to do this we would enter a value other than 0 here. But, in this case we want 0.
This is what the surface data window looks like now.
Next, we will do a similar thing for the location of the secondary mirror. We must keep the secondary mirror located at the back surface of the corrector lens. The THICKNESS of surface 2 needs to be the same as the THICKNESS of surface 3. But, you will notice that the signs are different, thickness 2 is 403.653mm while thickness 3 is -403.653mm.
This is why we select Minus thickness, so thickness 3 will be the opposite of thickness 2.
Click on the gray box under THICKNESS on surface 3. Move your mouse cursor over Pickup, then click on Minus thickness.
The pickup source surface is 2, so enter 2 and click OK.
Click OK to accept the default value of 0 for the pickup constant.
This is what your Surface Data window looks like now. Save this file.
You are now ready to set your variables and optimize this design. I will not show optimizing this design, but you can do so like you did on the first example. If you set radius 2 as a variable, then radius 4 will also vary and remain the same as radius 2. Likewise thickness 3 will vary when thickness 2 changes.
On the above Surface Data window I have highlighted the aperture radius of surface 6 (IMS). This is the semi-diameter of the field at the focus. This is where the eyepiece or film would be. So, 26.58mm is the semi-diameter for a semi-field angle of 0.5 degrees. In other words the size of a 1 degree field diameter is 53.17mm with this focal length of 3048mm.
Below, on the left is the lens drawing for the above design. On the right is the lens drawing for the original design before we added the special aperture data.
You will notice that surface 1 of the corrector lens now has a blank spot in the center. This is the special aperture data, which blocks the light so it does not enter the telescope. The drawing on the right does not show this blank spot.
You will also notice that the primary mirror now has a hole in it which the light would pass through. The drawing on the right does not have a hole in the primary mirror. Notice on the drawing on the right how OSLO shows the light passing right through the primary mirror as if it was not there. This is a case were OSLO does not "see" a design setup like we do. OSLO will only see a surface the second time if you enter it again at the proper location.
What I mean is that a non-perforated mirror (drawing on right) would reflect the light back toward the secondary mirror. If you wanted OSLO to show this additional reflection you would need to enter another mirror after the secondary mirror.
I have added the third mirror below. Its location is the same as the primary mirror's location, and its radius of curvature is the same also. This is surface 5.
Below is the lens drawing for the third reflection. The new focal length is shortened to 2128mm, which is f/10.64. The new focal position is about 146mm inside the primary mirror.
The setup with the central obstruction will have slightly different images and aberrations because the light near the axis is blocked. Here is the wavefront analysis which shows a hole in the center of the mirror.
Please contact me with questions or comments at: firstname.lastname@example.org