Please take a look on a software to help on Foucault data reduction , for Windows, just one .EXE  and one .INI file on the same directory to run, a text file is generated for each mirror. This data file  handles many measurements , showing  tabular or graphicaly, with one click , the progress of your mirror's profile, as long as action needed, along many figuring sections.
 
 

COUDER.EXE 

Sample Screens

Grinding an extra thin 40cm  (20mm thick)	Pulleys, belts and tiles plaster tools ...

I start grinding a plane glass using SiC grit 60 and a 50% size tile tool  (see it working at left picture ), Center over center, covering all mirror's surface. This procedure ensures that a good rough curve is being generate as saggita is increased.  When desired saggita minus 1mm is reached , almost all mirror's surface is ground, then a new 100% tool is made matching the  surface curve and used from grit 80  up to fine grinding with AlO grit 3000.

   When to change grits ? Simple. Look the mirror's surface with a magnifying lens and check if pits in the suface are about the same size. If there are larger pits, spend more time with current grit..

   At the end of fine grinding, mirror's surface shall reflect images at low incidence angles.

Another way to grind fast is use diamond tool. See some pictures how to adapt  a drill to the grinding machine above ( 1, 2, 3)

    The picture bellow looks like gigantic mirror , but in fact, unfortunately, is only a 10" pirex .
 
 

Reflections on a  fine grit surface

    Plate glass and hot  wheather are not a good combination, evaporation effects will cause an expected but minimizable TDE. Consider loosing up to 1" of your mirror under these conditions.
   I do all  polishing TOT (Tool On Top) with lots of weights over a plaster tool cast with the mirror curvature . I like to use polishing pads because of the speed to polishing .
   At the end the mirror surface is very close to spherical but surface is rough when seen  with foucault . A new tool is then prepared using picth and will be used to smooth the surface as well as figure to a parabola. Note the picture bellow when the CeO mix is at the point of best polishing effect, it´s very thin.
 

   I got better results with machine figuring than hand. The mirror on top , is hold by a round wood disk with four clamps . These clamps shall permit mirror to rotate in respect to the wood disk . That procedure always gives a really smooth surface without zones at all.
   The machine drive arm pin is placed on the slightly uncentered hole in wood disk , driving the mirror in a very complex movement over the tool and turntable.
 

Tool to hold mirror during polishing

    The folowing pictures shows many defects caused by machine polishing  BEFORE I begun to use the wood tool pictured above. The pictures from a 14" just below shows the drastic difference using this simple device.

   The pictures of Focault tests taken with Pentax  ME1, 150 mm lens , 8 sec exposition and then scanned. Unsharp Masking was used to enhance surface details . Focault with .5mm  pinhole:  Focault light is at right.

Mirror : 28 cm  diam, FL =132 cm , 20 mm thick tabletop glass

Original Image: ..	and after Unsharp Masking

After Parabolization 

Edge	70% zone	Center

    Enhanced by Unsharp Masking, pictures shows a small central bump , concentric rings are gone,  visible ripples are similar to pitch tool squares population,  5mm TDE visible as a very bright strip on picture at left . My Couder program indicates a 1/6 wave. I decided to stop because figuring a f/4.5 plate glass is very difficult. As one example , one minute figuring spell (one turn) changed error from 1 wave to 1/2 wave undercorrected, that's too fast to take good ontrol!

This mirror, after TDE masked, gave me good images up to about 300x , despite the ripples,  not bad for a net 11.5" and  $60,00 investment and much fun.
 

My recent  14"

Just after astigmatism removal	Images after machine parabolizing	using the MOT method mentioned	above.

Plot of Couder.exe of mirror above. Compare with images above. Max error = 1/4.3 wave at this point.

Saturn and Jupiter as seen by this mirror at this point (just brightness and contrast changed) and a B&W QuikCam on Newtonian focus :

     , visually I can see more definition, Saturn rings are 3-4 pixels wide only !

Some terrestrial pictures 1,  2,  and  3made also from this mirror .  People are 2.5 Km away .

After a period of  automatic machine smoothing and figuring using the strokes, the mirror profile is clearly improved and the smooth paraboloidal shape is finally there . See picture below and note is the same mirror  .  Note a kidney shaped defect  as a shallow depression area that appears on pictures above at 1:00 on "Just after astig..."  and at 11:00 on "using  MOT...". This same region is at 7:00 on pic bellow. This mirror has a 1.3 cm ring of TDE to be masked.

Results with submerged polishing with syntetic ferric oxide :

  ****   More on ATM in portuguese ****
 
 

  The support for the 40cm mirror (16"), 20 mm thick is a design by Richard  Schwartz  , from Star Fleet Engineering  using Finite Element Analisys .  The result is a maximum flexure of 1/20 wave with 12 point support.

Telescope Making Web Ring

Next | Previous | Random | List Sites Next 5 | Previous 5 | Join