Planning two achromats - no melt info
Posted 14 March 2013 - 12:27 AM
I would like to make two f/15 achromats, both based on formulas given in James Baker's 1963 article in Applied Optics - one a Fraunhofer type, and the other a R2=R3 formulation. However, while I am reasonably certain what kind of glass I have, I do not have specific melt info. I do have catalog RI and partial dispersion values from Bob May's web site.
I plan to make the R2=R3 lens first, and will make accurate test plates for R1 and R4. Also before I get started on the lenses I will make an 8" flat (or three!) for testing by auto-collimation (I'm also planning on making a 10" plano convex lens from 3/4" plate glass as a collimator for fringe testing). I have a 546nm narrow band filter and will obtain 486nm and 656nm filters before I get to the lens testing stage. So I should be able to measure and plot the actual correction of the lens for comparison to OSLO design plots.
I will also make and measure a set of aluminum screw-together rods so that I can measure the distance from glass to knife edge within about .005" - .008"
Does anyone have a suggestion as to the expected variance of glass refractive indices around the catalog values for old B&L glass like this? My thought is to calculate the lenses with the catalog values, and the to calculate what the plots would look like if either or both pieces of glass had higher or lower RI values by some typical amount. For instance, NBK-7 in the catalog has a N-d refractive index of 1.51680, while the 9.25" blank that JIMEGGER described in his 9/20/11 post had melt data index of 5.1711, a difference of 0.0031. Is this a small variation from the catalog value? typical? larger than typical? Would you expect old B&L glass to vary more than modern Schott glass?
This will be my first work on lenses, but I have made four mirrors.
I realize that I've gotten down into the weeds on this, and that I might do fine just making a lens using the catalog values and then doing corrections afterwards. However, it is so easy to plug different glass values into OSLO that making an "envelope" of six plot variations around the expected plot seems like a good preparation.
Posted 14 March 2013 - 09:43 AM
What you can do is make the crown element first since it is positive lens. Since you know the two radii and thickness, measure the exact focal length in monochrome light. You can do it at a couple of wavelengths and also stop the lens down since it will show spherical aberration. Knowning the measured focal lengths and the calculated ones you can see how far off they are and that will tell you how different the refractive index of the glass is.
Once both elements are finished and the lens assembled you can again measure the positions of were the green, red and blue wavelengths come to focus and check them against the OLSO data and adjust the refractive index in OLSO to match the focal lengths you measured.
Posted 14 March 2013 - 11:02 AM
The following numbers are from the ATM III:
R1= .5814 cvx
R2= .3585 cvx
R3= .3585 ccv
R4= 1.6189 cvx
Multiply your desired FL by the above factors and you are all set!
From Baker's Telescopes and Accessories, the Franuhofer design:
R1= .6083 cvx
R2= .3544 cvx
R3= .3585 ccv
R4= 1.4936 cvx
Airspace: Shim just enough to prevent central contact.
I could give you thicknesses, but at f/15, variations in thickness don't count for that much.
Now a lot of theoreticians will have a lot to say about refinements of the design, and make you fuss over melt data and thicknesses but the numbers here are surely enough for what you want.
Posted 14 March 2013 - 03:19 PM
Posted 14 March 2013 - 10:48 PM
With respect to your suggestion for "stopping down" the crown element for testing, I was thinking of making a set of zonal mask so I could plot the zonal aberation similarly to the plot in OSLO, with difference in mm from the N-d central back focus on the x axis and lens radius on the y axis.
This brings up a question I have about one of the recommended testing proceedures in the Remer book, the "modified Hartmann test". I can't see how it adds any information to what I can get from a knife edge tester with narrow band filters and zonal mask. Am I missing something?
Ed - Thanks for the specs - I will have the opportunity to look at the original source for the Baker lens in a couple of days - two days ago I ordered a "very good" copy of the Telescopes and Accessories book through Amazon for $4.00 plus shipping - what a find! I appreciate that thickness will not make much difference, but I will probably have a go at trying to optomize for somewhat thicker lenses since I will need to cut the blanks down to thickness before generating curves - I thought I would aim for about 0.45" for the crown and 0.40 for the flint - is that about right?
Thanks for all comments.
Posted 15 March 2013 - 08:54 AM
Attached is an OSLO file for the Baker R2=R3 design using the data from 'Telescope and Accessories' and the RI data for Bob May's site for BSC-2 and DF2. Using the "SCALE LENS" function in OLSO, enter the focal length of the lens you want and OSLO will calculate the design. Just be sure to adjust the Aperture radius to match your lens. You can then easily check what happens if the RI index is different or how sensitive the thickness of each elements is to the performance of the lens.
Posted 15 March 2013 - 09:11 AM