Jump to content

  •  

CNers have asked about a donation box for Cloudy Nights over the years, so here you go. Donation is not required by any means, so please enjoy your stay.

Photo

Testing a Diaganol Mirror

  • Please log in to reply
14 replies to this topic

#1 steve t

steve t

    Viking 1

  • *****
  • topic starter
  • Posts: 590
  • Joined: 17 Jun 2009
  • Loc: SW Ohio

Posted 13 February 2020 - 12:30 PM

Hi All,

I've taken my first tentative steps into testing optics by purchased a 1" flat from Surplus Shed. (Thanks Dave for the heads up)

 

My goal was to test two 1.00' diagonals, to see if one was any better than the other. Both mirrors were purchased from Newport Glass and are stamped, on their backs, as being 1/10 wave. 

 

I used a CFL light with the diagonal mirrors laying flat on a table.

 

I placed the flat on the diagonal and looking down through it I had trouble seeing any bands. Applying a very slight pressure to the flat, the bands were faintly seen.The bands on both diagonals looked straight, but were so faint I couldn't photograph of them.

 

Does anyone have any suggestions on my setup and how to make the bands more prominent so they could be photographed? 

 

Thanks

Steve T


  • vtornado likes this

#2 Chuck Hards

Chuck Hards

    You don't know Swift from Astrola

  • *****
  • Administrators
  • Posts: 26,710
  • Joined: 03 May 2010

Posted 13 February 2020 - 12:36 PM

The bands will be faint under fluorescent light.  They will be more easily seen with monochromatic light.    

 

But if you saw straight, parallel bands, even faintly, you have a good diagonal.


  • steve t likes this

#3 steve t

steve t

    Viking 1

  • *****
  • topic starter
  • Posts: 590
  • Joined: 17 Jun 2009
  • Loc: SW Ohio

Posted 13 February 2020 - 12:52 PM

Chuck,

Thanks for the Information.

 

For a monochromatic light source, I may try a red led flashlight next.

 

Steve T. 



#4 Chuck Hards

Chuck Hards

    You don't know Swift from Astrola

  • *****
  • Administrators
  • Posts: 26,710
  • Joined: 03 May 2010

Posted 13 February 2020 - 12:54 PM

Green light is most commonly used in optical shops, can you find a green LED?   



#5 DAVIDG

DAVIDG

    Cosmos

  • *****
  • Posts: 9,245
  • Joined: 02 Dec 2004
  • Loc: Hockessin, De

Posted 13 February 2020 - 01:31 PM

 A LED is not monochrome. It has band width of around 50nm. You need a light source with a band width of an angstroms or less  ( 1nm = 10 angstroms)  The CFL bulbs works because of the gases inside the bulb produce a number of monochrome emission of different wavelengths but they are  discreet and very narrow. The white coating in bulb partially diffuses them  to make the light look white but not enough that the light will still produce interference fringes of different colors.   That is one reason why the fringes your see are low in contrast because there are a numbers of them of different colors right next to each other.  The other issue is that if the diagonal your testing is aluminized.  There is large difference in reflectivity between that surface and the bare glass surface of the reference flat. 

   So what can you do to increase the contrast ?  First you want  test with a  monochrome source. The simplest way to do that is to filter the light from the CFL bulb so your only viewing one wavelength. That is simple to do. Just look thru a green, red or yellow filter when your doing the test. The gas emission in the  CFL bulb are far enough way from each other in wavelength that a simple filter, of  the color you want to use will isolate that wavelength from the others. I have used clear red or green candy wrappers to look thru to increase contrast.

   The second is harder to do. That is to get a better match in reflectivity between the aluminized surface and the bare glass. That requires the surface on the reference flat be partially coated to increase it's reflectivity.  I have flats that are coated for this. 

The good news is that all you need to do is see the fringes, even faintly to judge the flatness. What is critical thou is the surfaces are very clean and that when you see the fringes, you want no more then 5 of them being visible. This  greatly increases the sensitivity of the test. When more are visible it is hard to judge how much of any they depart from straight and when many are visible a surface that is waves from flat will show straight fringes.

 

                               - Dave 


  • steve t and jcruse64 like this

#6 steve t

steve t

    Viking 1

  • *****
  • topic starter
  • Posts: 590
  • Joined: 17 Jun 2009
  • Loc: SW Ohio

Posted 13 February 2020 - 01:55 PM

Thanks Dave,

I'll dig around to see if I got any clear green wrapping material.

What I saw were many faint lines close together, so as you said I don't think my first try was very sensitive.

Steve



#7 DAVIDG

DAVIDG

    Cosmos

  • *****
  • Posts: 9,245
  • Joined: 02 Dec 2004
  • Loc: Hockessin, De

Posted 13 February 2020 - 02:08 PM

 Steve,

    Here is basically what you should see. This is with my monochrome light source and both  the reference flat  and the diagonal I'm testing are not coated so both of these factors greatly increase the contrast.  This picture shows a few more fringes then it should to get a real good estimate of how flat the surfaces is, thou. 

 

             - Dave 

 

coulterdiagonalafter.jpg


Edited by DAVIDG, 13 February 2020 - 03:03 PM.

  • steve t and Bomber Bob like this

#8 steve t

steve t

    Viking 1

  • *****
  • topic starter
  • Posts: 590
  • Joined: 17 Jun 2009
  • Loc: SW Ohio

Posted 13 February 2020 - 02:51 PM

Dave,

Thanks for the photo, it helps me picture what I'm looking for. The diagonals I'm trying to test are coated.

 

I took a small carboard shipping box and cut a 1" hole in the top. This allowed the light from the CFL shine on the bottom. I then put a dark green eyepiece filter over the hole so I would get a green spot a little larger than the diagonal. I also defused the light by placing tissue paper over the green filter. I cleaned both surfaces before testing.

 

Still no luck seeing the bands. Once and a while I could just make out a bulls eye pattern, but it was very faint.

 

I just happen to think, are both surfaces on the flat polished to the same level of accuracy?\

 

Thanks

Steve T



#9 DAVIDG

DAVIDG

    Cosmos

  • *****
  • Posts: 9,245
  • Joined: 02 Dec 2004
  • Loc: Hockessin, De

Posted 13 February 2020 - 03:20 PM

 Steve,

   The diagonal your testing goes on the bottom. The reference flat which is clear is place on top  of the diagonal and the light from the CFL bulb  shines down on the assembly like my picture. The surfaces must be extremely clean. A piece of dust will hold them part and you either see no fringes or many of them.  You can press lightly on them to see if the fringes become visible and also reduce their number. 

   You don't have to filter the light coming the CFL bulb itself. All you need to do is look through the filter  and at the flats. It also doesn't need to be dark green filter. The green emission from a CFL bulb is from Mercury and it is at 546nm which is more of a  lime green like in my photo.  There is also a  red emission from hydrogen  and yellow emission from sodium. So try those color as well and see what gives the best contrast. By the way if see a bullseye, that means one the surfaces is far from flat.

  In the good old days you made a monochrome light  by burning table salt in  a candle flame. It burns bright yellow and is monochrome. so you would see bright yellow and black interference fringes.

You might want to try just using your surplus shed reference flat against another piece of uncoated  "flat"  glass. like a table top or a  piece of glass in a picture  The fringes should be much easier to see and they most likely will be all twisted up since window type glass is not very optically flat. Then you can get idea of what to expect.

 

                        - Dave 


  • steve t likes this

#10 steve t

steve t

    Viking 1

  • *****
  • topic starter
  • Posts: 590
  • Joined: 17 Jun 2009
  • Loc: SW Ohio

Posted 13 February 2020 - 03:47 PM

My set up is as you described.

Instead of filtering the light coming in I'll try looking through the filter. I'll see what other filters I got.

Prior to testing I cleaned the surfaces using First Light optical cleaning method. 

Thanks again.

Steve T. 



#11 steve t

steve t

    Viking 1

  • *****
  • topic starter
  • Posts: 590
  • Joined: 17 Jun 2009
  • Loc: SW Ohio

Posted 13 February 2020 - 04:17 PM

Dave,

Success, holding a lime green filter in front of my eye did the trick.

 

I wasn't able to get the number of bands to less than about 10 without applying a lot of pressure on the flat. The bands were smooth, but with a slight bow to them. I need to do some research to see if the results can be quantify.

 

I'll try to get some photographs posted later.

 

Steve T



#12 DAVIDG

DAVIDG

    Cosmos

  • *****
  • Posts: 9,245
  • Joined: 02 Dec 2004
  • Loc: Hockessin, De

Posted 13 February 2020 - 05:06 PM

 Steve,

   The surface is dirty  if you can only get to ten fringes  It only takes one piece of dust to hold them apart.  The bowing  of the fringes is only going to get worse with fewer bands.

 Here is how to calculate how flat the surface is, first  rule is that the spacing between the fringes is always 1/2 wave, no matter if you have 4 shown or 44 shown. You only want a few fringes showing because they will be far apart and so you have better accuracy in measuring how far apart they are. Now you draw a  line down one of the fringes from were it starts to were it ends and you measure the maximum deviation from the straight line. To find  how flat the surface is you take the ratio of the deviation from straightness divided by the spacing between the fringes . Since the spacing between the fringes is always 1/2 wave you divide that result by 2 to get the result in waves vs 1/2 waves

 So say for  example you measure the spacing between the fringes as  0.5" and the fringes are bowed so the when you draw a straight line down from the start to end the max deviation  is 0.25"  ( ie 1/2 of the spacing between the fringes). The flatness is (0.25/0.5)  /2 = 0.25 ie 1/4 wave. 

   So if your diagonals are flat to 1/10 wave, the deviation of the fringes from straight needs to 1/5 of the distance between each fringe. So they should look pretty straight. 

 

 

                           - Dave 

     flatness calculation.jpg


  • steve t likes this

#13 steve t

steve t

    Viking 1

  • *****
  • topic starter
  • Posts: 590
  • Joined: 17 Jun 2009
  • Loc: SW Ohio

Posted 13 February 2020 - 05:44 PM

Dave,

I cleaned both again and it's better. Since the flat is smaller than the diagonal do you think it would be better to use a larger flat?

Steve T.



#14 steve t

steve t

    Viking 1

  • *****
  • topic starter
  • Posts: 590
  • Joined: 17 Jun 2009
  • Loc: SW Ohio

Posted 14 February 2020 - 10:08 AM

I was able to get the surfaces clean so I was able to get the number of fringes down to four.

Keeping in mind that the edges of the fringes being very diffused and I did the conversion correctlysmile.gif 

 

I measured the distance between the fringes to be about 4mm.and the bow in one of the fringes to be about .5mm. Using these numbers and the equation, provided by Dave above, I came up with a surface of 1/16 wave over the 1"  area that I was able to test at one time.

 

Again the bow in the fringe was very difficult to measure because of it diffused edge.

 

Since I've started down the dark side of optics testingsmile.gif . I still may look for a larger flat that will allow me to test the whole surface at one time.

 

Steve T



#15 davidc135

davidc135

    Viking 1

  • -----
  • Posts: 777
  • Joined: 28 May 2014
  • Loc: Wales, UK

Posted 14 February 2020 - 11:58 AM

Dave, In your illustration shouldn't a fringe or half wave be measured from the inside of one fringe to the inside (not the outside) of the neighbouring fringe? So the blue line should be longer.  David


  • steve t likes this


CNers have asked about a donation box for Cloudy Nights over the years, so here you go. Donation is not required by any means, so please enjoy your stay.


Recent Topics






Cloudy Nights LLC
Cloudy Nights Sponsor: Astronomics