Making an Optical Flat Mirror
One does not hear about making flat mirror in ATM too often. It seems that even people making their own primary mirrors contend with buying diagonal mirror from commercial sources. Part of it may be that making a good diagonal is a more demanding job that grinding and polishing a lot larger paraboloid mirror which it will be used with - draw your own conclusions from it :-)
Description below is intended for those interested in persuing diagonal building though it is worth pointing out that flat optical surface has other applications as well. Large optical flat mirror can be used as as coelestat in solar telescope. It may constitute back side of corrector plate. As another example, objective prism contains two flat surfaces.
I have build my first diagonal during construction of a 6 inch f4 Dobsonian. Description below lists details and progress of making large diagonal mirror (5 inch minor axis) for 24 inch Dobsonian under construction by Kingston Centre - RASC.
Technique I am using is derived from usually scant descriptions found in a relatively few old ATM books.
Albert G. Ingalls (ed.) Amateur Telescope Making, Scientific Americam Inc. 1926, 1928, 1933 &1949
Allyn J. Thompson, Making Your Own Telescope, Sky Publishing Corp. 1947
I ended up with a combination of those methods based on personal preference and accumulating experience.
One is not enough ...
The paradox of making a flat mirror is that they like company. Actually, the easiest and least frustrating way is to make three of them at the same time. Why?
So what to do with extra flats - afterall you are probably building just one scope. Or aren't you :-) ? I ended up with using extra one in a 20 inch Dobsonian.
Step 1. Grinding
I have started with three pyrex disks - each 7 inch in diameter. I have marked them as A, B and C.
Those blanks were then ground with #400 carborundum in a following order: A on B, B on A, C on B, B on C, C on A and A on C. Then the second cycle started with B on A, A on B, B on C etc. And finally the first cycle was repeated. Carborundum was charged only once for each disk combination (6 times in each cycle) and each charge was ground with quater stroke and some pressure for three circles around the stand. All that took about 2 hours.
Images below show the appearence of one disk prior and after #400 grinding.
original 7 inch Pyrex blank
flat surface after #400 grit
Then A, B and C mirrors were groud in the same alternating order using #600 carborundum (3 cycles) and finally #1200 aluminum oxide (3 cycles).
Step 2. Polishing
Polishing was done on a simple turntable machine powered by a drill press (through a set of reducing pulleys). Tool was made of pitch as hard as I could manage to pour and before each polishing session it was cold pressed for a few hours with lots of weight (do not even think about warm pressing for polishing flat mirror).
Borosilicate glass is hard and machine polish is slow so it took about 100 hours per surface (x 3 !) to finish polishing.
I have started to test mirrors by observing interference fringes after about 20 hours of polishing. Red laser pointer was used as a source of monochromatic light - of course it had to be diffused (I have used a number of layers of paper tissue). Before testing, it is important to wait at least 1 hour after finishing polishing session and make sure the room temperature did not change in a meantime!
Calculating the surface of each mirror is described in books mentioned above. When the surfaces were completely polished, I figured them by hand on a turntable by polishing over different zones as needed. I had to use a lot of preasure to slowly deform each surface toward flat.
In the end all three surfaces were better than 1/8 wave across 7 inch diameter.
Step 3. Cutting elliptical mirror
Now the scarry part - had to cut elliptical diagonal out of each polished flat!
First, single flat (with optical side clearly marked!) as well as pieces of window plate glass (6 mm thick and slightly larger) were slowly wormed up in an oven. Then bee wax was melted on optical surface as well as on one of the other pieces and all three were sandwiched together. Air bubbles were squeezed out and the "sandwich" was left to slowly cool down in the oven.
Next, that "sandwich" was locked tilted at 45° inside block of plaster. It is important have optical surface facing up - othervise one side WILL BE broken off!
Big hole drill was constructed out of 4.5 inch diameter block of wood and metal pipe made by soldering piece of steel sheet. Drill was powered by set of belts and a worm gear reducer to run at about 200 rpms.
Carborundum #80 was fead as water slurry during "drilling". It took 2 full days to cut out each 4.5 inch minor axis diagonal. The depth of cut was about 7 inches!
4 inch diagonal, on the other side took only one day. That's a lot less noise!
All three diagonal flats were then coated with enhanced aluminum at Moonward Vacuum Coatings in Sudbury, Ontario. Diagonal shown below (4 inch) is already mounted in my 20 inch scope. 4.5 inch diagonal went into 24 inch Venor Scope built by members of Kingston Centre of RASC and the second 4.5 inch is left over for future Centre projects.
Back to 20 inch scope
Number of visitors:
© Jan Wisniewski