971 replies to this topic

[DAVIDG]

 Ok time to see if I can clean up the Schmidt corrector in my Dynamax 8. So I'll be posting what I do and why I'm doing it. I'm not  a Dynamax hater or an optical snob. My purpose for my posts are to try to educate others so they have the information they need to make  an educated judgement of their optics and in this thread use the information to test and possibly refigure their own corrector 

   So the first step  was to pull the corrector and remove the secondary assembly. I remove the screws around the retainer ring and carefully lefted the corrector out. I then removed the three screws on the back of  the secondary housing to gain access to collimation screws for secondary. I then removed them while holding the secondary with piece of lens tissue.  With the secondary out of the housing I unscrewed  the housing  which is in two pieces and removed it from the corrector plate.

 

                - Dave

 

Edited by DAVIDG, 30 January 2017 - 09:20 AM.

[DAVIDG]

  With the secondary out I noticed the hole in the corrector were the secondary is mounted had a number of chips around the edges and the edges were sharp.  The edges needed to be cleaned up since if a microchip of glass broke off when I was polishing the glass it would scratch up the surface. So I used  a wet stone rod to grind the edge smooth on both sides while running water over the surface, Here are the before and after pictures.

 

                           - Dave 

 

 

Edited by DAVIDG, 29 January 2017 - 09:42 PM.

[DAVIDG]

 The next step was to test both sides of the corrector and  see what the figures looked like. To do this I placed each side of the corrector in contact with an 8" optical flat I have. I got this flat off of Ebay for $35 because it has a few minor scratches which don't effect it at all for testing purpose. When I tested it  I was very please to see that it is flat to at least 1/10 wave across the complete surface and when new I'm sure it cost at least $500.

   Schmidt correctors can have one side "flat" and the other with the Schmidt curve or both side with the Schmidt curve but with 1/2 the power needed on each side. Criterion put all the correction on one side. The flat side doesn't needs to be  optically flat. It can be a few waves from flat and it doesn't do any harm but it needs to be optically smooth. So what you want to see are  ideally straight interference fringes or at least smooth arcs which indicated either a slightly concave or convex surface. What I saw was band of widely spaced and asymmetrical fringes down the middle of the plate. This indicated that the middle  section  was closer to the surface  of  the optical flat. So the shape of the back surface is optically rough, convex cylinder. Here is a pictures that shows the fringes.

 

                    - Dave  

 

Edited by DAVIDG, 30 January 2017 - 09:42 AM.

[DAVIDG]

So the next step was to test the surface that has the Schmidt Curve. What you want to see is a set of fringes that are evenly spaced  and smooth  that define the neutral zone of the plate. The neutral zone should be round and not oval or have any asymmetry to it. The neutral zone is the deepest part of the Schmidt curve and in the exact center of it the light goes straight thru it with no deviation.

   What I saw were 4 areas at about 90 degrees  apart were the neutral zone had high areas. You can see them at the 12, 3, 6 and 9 o'clock positions on the plate. Also the neutral zone wasn't around but slightly oval shaped. The Schmidt curve is asymmetrical and causing the astigmatism I was see when I tested the complete system via double pass autocollimation. Here is a picture that shows the problems.

 

     - Dave 

Edited by DAVIDG, 30 January 2017 - 09:22 AM.

[DAVIDG]

 The next step was to cast a pitch lap on an 8" mirror blank and press it against the 8" optical flat. The lap will then be used to polish the flat side of the corrector to clean it up. One of the "rules" in figuring optics is the glass follows the shape of the lap and the  lap then conforms to the shape of the glass. So by pressing the lap against the optical flat the pitch will be as flat as it's surface. Then I  will polish the flat side of the corrector using this pitch lap. So in theory the corrector will polish flat as well. Ever 10 minutes or so I'll repress the pitch lap against the optical flat to keep it flat. I used my gas grill on my deck to melt the pitch and cast the lap. Here are pictures of the 8" mirror blank with  a dam of tape to hold the melted pitch and the finished lap ready to start polishing the corrector with. So the next step for the next couple of days will be to polish  the flat surface of the corrector  and see what happens. Stay tune for more results ! 

 

             - Dave 

 

                      

Attached Thumbnails

• 

Edited by DAVIDG, 30 January 2017 - 09:23 AM.

[rcwolpert]

This is REALLY interesting!

[Mitrovarr]

It would be kind of annoying if it wasn't the shaped side of the corrector that causes the problems, but the flat side. I can imagine it being difficult to make the curved side, but messing up the flat side seems like pure carelessness.

[petert913]

This is fascinating !  I can't wait to follow more of this project.

[deSitter]

This is fantastic!! Thank you!

 

-drl

[highfnum]

wow nice lesson

it verifies  suspicions

but hard facts are the truth makers

[DAVIDG]

 I put in about 45 minutes of polishing on the "flat" side. It's starting to clean up. The big high zone down the middle of corrector is starting to smooth out but I have a couple of more hours to go before it will be smooth enough for me. Here is a picture so one can compare to what it looked like when I started.

 I'll see  if can find  some time tomorrow evening to continue the work.

 

                    - Dave

 

[starman876]

this is great information.  I take it your doing this by hand and you dont have a grinding machine?  

[highfnum]

Oh Dave are you using polish agent in pitch like cerium oxide

[DAVIDG]

 I'm doing this by hand sitting at my kitchen table  while watching  TV. I'm using a small amount of cerium oxide and water on the pitch lap. 

   

                      - Dave 

[Jeff B]

So cool Dave.

[tim53]

I'd like to hear more about that cool light panel!

[highfnum]

are you using short strokes in a random circle pattern?

[Terra Nova]

Wow! Thank you so much! I am learning a lot here!!

[DAVIDG]

 Thanks everyone who "like" this thread. I have 70 "likes" so far !  Please note that I am posting images of the test results so everyone can see what I'm talking about vs saying that because I have been observing for 35 years I know good optics from bad and  because of this I can  stating the quality of the optics I'm using.

   The stroke I'm using  to polish the back surface against the pitch lap is the  classic "W" with about 1/2" of the corrector over hanging the pitch lap as I push the  corrector  over the lap in making a 'W". The green light panel is an UniLamp, Edmund sells them  for $425  dollars http://www.edmundopt...tic-lamps/1659/ but again Ebay to the  rescue.  I got a couple of them and paid under $30 each. You can make your own. All it is are two uncoated UV fluorescent bulbs. There is Mercury in the bulbs and it has a very strong emission at 546nm which is bright green. There is a  piece of green plexiglass to isolate the green emission from the other wavelengths  and a white piece of plexiglass to act as a diffuser. You can also use a cheap CFL bulb.  

 

              - Dave 

Edited by DAVIDG, 30 January 2017 - 09:51 AM.

[starman876]

Now I have to ask this. Why green?

[DAVIDG]

Now I have to ask this. Why green?

 It doesn't have to be green,  just monochrome. It is just that fluorescent lamps use Mecurcy which has a number of emission lines one of which is the green line which is very bright and also there are no other emission lines close by it. So it is very easy to isolate just the the green one with a wide band green filter like a piece of green plastic. In the old days they would us  an alcohol lamp and coat the wick with salt. When sodium burns it gives off a bright  monochrome yellow light. You  can also use a sodium street light and a yellow filter to do the same. 

   The shorter the wavelength thou the mores sensitive the test becomes. This is because the spacing bewteen the fringes is always 1/2 of the wavelength of light used  So if you used a HeNe laser which operates at 636nm vs the Mercury source at 546, the Mecurcy source is more sensitive. People have used Helium lamps since there is 501nm emission that one can isolate fairly easily and pick up more sensitivity. For what I'm doing the Mercury source has plenty of sensitivity thou. 

 

                    - Dave 

[starman876]

So I guess from the wavelength used and how the interference lines show up you measure the difference between the lines to come up with the flatness?  I take it that is all dependent on the accuracy of the the flat.  In my line of work normally the calibration device has to be 10 times more accurate for the test results to be of any value.  I take it that kind of accuracy is not needed here?

[DAVIDG]

So I guess from the wavelength used and how the interference lines show up you measure the difference between the lines to come up with the flatness?  I take it that is all dependent on the accuracy of the the flat.  In my line of work normally the calibration device has to be 10 times more accurate for the test results to be of any value.  I take it that kind of accuracy is not needed here?

  I'm not trying to get the back surface flat to 1/10 wave. I'm trying to get it optically smooth.  I know the flat is flat and smooth since I tested it against other flats I have. What I want to see are uniformly spaced lines that are ideally straight  across the whole surface or  at least smooth arcs. Again that are uniform and evenly spaced.  If the lines were straight that means that the two surfaces have the exact same shape. If they are  smooth arcs then the one surface is spherical vs the other. As you can see from the pictures they are a mess  so nothing close to optically smooth. 

  When it comes to measuring flatness the spacing between the fringes is always 1/2 a wavelength of light. So you measure the distance between two fringes and then draw a straight line down the fringe and measure the maximum deviation from straightness.  You can the sensitive by  expanding the fringes by lightly pressing on the surface. The ratio of those two values tells you how much they deviate from each other. Knowing that one surface is flat to some amount tells you what fraction of the wavelength or multiples if it far from flat they differs from each other.  Any of the classic telescope making books goes into detail on how and why it works. I don't want to go into more detail because I don't want this thread to drift from it's purpose.

 

                - Dave  

[SteveGR]

You're a braver man than I.  But this is fascinating, thanks for posting it.  Very curious how well it comes out.  Was the corrector the limiting factor?  Are the primary and secondary of good quality?

[DAVIDG]

You're a braver man than I.  But this is fascinating, thanks for posting it.  Very curious how well it comes out.  Was the corrector the limiting factor?  Are the primary and secondary of good quality?

 I tested the complete system using double pass autocollimation and the primary by itself via Foucault and Ronchi at the radius of curvature. The primary was a perfect sphere. The issues is the corrector. You can see the test results here http://www.cloudynig...g-of-dynamax-8/  The corrector has a fair amount of problems so I doubt that I could make worse and since I have refigured Schmidts and many other system before   I'm pretty sure I can make this one right. It is just going to take some time. Once the corrector has been refigured any residue errors in the system will be removed by refiguring the secondary until the complete system nulls. 

 

                   - Dave 

Edited by DAVIDG, 30 January 2017 - 01:56 PM.