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

Turning a spherical mirror into a parabolic mirror

This topic has been archived. This means that you cannot reply to this topic.
22 replies to this topic

#1 Orion64

Orion64

    Apollo

  • -----
  • topic starter
  • Posts: 1155
  • Joined: 29 Nov 2009

Posted 03 July 2014 - 06:33 PM

I have seen people mention that when they grind mirrors, they first grind and polish it off into a spherical mirror and then do the last polishing to form the parabolic shape. Can this be done in this way? Why not just form a parabola in the first place?

#2 Noisykids

Noisykids

    Apollo

  • *****
  • Posts: 1100
  • Joined: 01 Jun 2011

Posted 03 July 2014 - 06:52 PM

in my experience from grinding and polishing a 6" f8, and from reading a lot about it, the natural random motions of grinding and polishing two pieces of glass together will naturally result in a spherical mirror. to make it a parabolic mirror involves, i read, not very much more polishing, a couple of minutes maybe, if done in the right motions. a spherical mirror has a constant radius of curvature, but not a precise point of focus. a parabolic mirror has it the other way around. the radius of curvature is not constant, but the point of focus of the reflected light is.
i tested my mirror on a homemade foucalt tester and as i dropped the straightedge across the light beam the mirror went completely and evenly gray. from what i understood at the time, this was a good thing. so i brought it to work to show it to a buddy, and he knocked it off the test stand onto a stainless steel counter and made a big ding in the edge. so i didn't parabolize it and it was fine. i think parabolizing a mirror like that involves polishing off millionths of an inch of glass. that end point would be kind of difficult to get to without going for the sphere first. believe me, i'm no expert at this, i only made the one mirror. but i think that's the gist of it.

#3 gpelf

gpelf

    Ranger 4

  • -----
  • Posts: 331
  • Joined: 28 Dec 2011

Posted 03 July 2014 - 07:48 PM

+1 a sphere is easy to obtain and a parabola is not far behind. On slow F ratio mirrors 10min of polishing can reach the needed change. On a fast mirror the changes involve would be very hard to obtain without the sphere.
Greg

#4 gregj888

gregj888

    Vanguard

  • -----
  • Posts: 2249
  • Joined: 26 Mar 2006

Posted 03 July 2014 - 08:39 PM

+1, Brett, it depends on the size and focal length of the mirror.

My 6" half way to a parabola for a trischiefspiegler (don't worry about it) took 11 "W" strokes to figure. My 20" f 3.5 has taken hours on a machine and still isn't done.

The 20" could have been ground and polished to a parabola, though that too is tricky. So... depends.

#5 Pinbout

Pinbout

    ISS

  • *****
  • Posts: 22761
  • Joined: 22 Feb 2010

Posted 03 July 2014 - 08:57 PM

Why not just form a parabola in the first place?



because the strokes are different. grinding and polishing are strokes that form a sphere. its simple and you don't have to think about the stroke. Even polishing is kind of a higher level caveman work.

parabolizing is totally different. Slower thoughtful strokes with specific intent.

#6 gpelf

gpelf

    Ranger 4

  • -----
  • Posts: 331
  • Joined: 28 Dec 2011

Posted 03 July 2014 - 09:02 PM

Well said Danny
Greg

#7 MKV

MKV

    Cosmos

  • *****
  • Posts: 7989
  • Joined: 20 Jan 2011

Posted 03 July 2014 - 11:47 PM

Why not just form a parabola in the first place?



because the strokes are different. grinding and polishing are strokes that form a sphere. its simple and you don't have to think about the stroke. Even polishing is kind of a higher level caveman work.

parabolizing is totally different. Slower thoughtful strokes with specific intent.

Danny, a 20" f/3.5 paraboloid and a sphere differ by 21 waves. Good luck parabolizing (figuring) that!

Mladen

#8 kfrederick

kfrederick

    Soyuz

  • *****
  • Posts: 3996
  • Joined: 01 Feb 2008

Posted 04 July 2014 - 01:56 AM

Correcting a 20 f 3.5 is not hard from a spherical mirror That is how I did this 36 inch f 3.5 the dark spots are fingerprints

Attached Thumbnails

  • 6615865-006.JPG


#9 kfrederick

kfrederick

    Soyuz

  • *****
  • Posts: 3996
  • Joined: 01 Feb 2008

Posted 04 July 2014 - 02:14 AM

The way I like to make a mirror is if you have a turntable is to spin grind and polish [no stroke] the mirror until fully polished . It will come off this close to spherical Then stroke polish the correction in .by taking the canter down and shorten the final RC

#10 Orion64

Orion64

    Apollo

  • -----
  • topic starter
  • Posts: 1155
  • Joined: 29 Nov 2009

Posted 04 July 2014 - 03:35 AM

Thanks for all the replies. So a 6 inch f/10 spherical mirror should be a parabola within a few minutes of polishing?

#11 MKV

MKV

    Cosmos

  • *****
  • Posts: 7989
  • Joined: 20 Jan 2011

Posted 04 July 2014 - 05:02 AM

Correcting a 20 f 3.5 is a piece of cake from a spherical mirror that is how it is done . That is how I did this 36 inch f 3.5 the dark spots are fingerprints

Great! That's 38 waves of correction. How long did it take you to do that?

#12 Mark Harry

Mark Harry

    Cosmos

  • *****
  • Posts: 8121
  • Joined: 05 Sep 2005

Posted 04 July 2014 - 05:36 AM

A 6" F/10 can be finished up from a sphere in as little as 10 minutes with thin compound. I've made dozens of these that rate with extremely minimal error, and which have performed essentially perfectly for customers.
They vary from a sphere only a tiny amount, so shooting for a sphere gets the maker very close to being finished. He also learns what strokes can get to this point of being close to done. My take, if the tyro doesn't know how to automatically get a nice sphere without an edge problem, the chance of making a parabola from the outset is very small.
M.

#13 KerryR

KerryR

    Soyuz

  • *****
  • Posts: 3796
  • Joined: 05 Dec 2007

Posted 04 July 2014 - 10:54 AM

Having made a 6" f11 sphere as my first mirror, I can tell you that getting a good sphere with no turned down edge is NOT easy if you're a beginner, at least not at that focal length. The learning curve if you're a complete beginner with no outside assistance (other than forums) is very steep. It took me several efforts, with each effort getting easier. Accidentally turning the edge is very easy with shallow curves, and can take quite a bit of effort to fix when you don't really know what strokes out of the many possibilities would be most effective, something for which you need either tons of experience from experimentation or expert guidance.

My next mirror was a 6" f5. I got the sphere and good edge much more quickly, but not automatically, and the paraboloid took several attempts before I got a smooth reasonably well corrected surface. Foucault testing contains it's own mysteries and idiosyncrasies for an un-guided neophyte as well.

Perhaps an f8 or f10 sphere is easier to get, but I think one of the great ATM myths may be that spherical surfaces are automatic. I think close spheres are almost automatic, if you pay attention to grinding and polishing strokes, but a good, smooth sphere with no hint of TDE is not automatic for most beginners (or in my case, second-timers).

#14 kfrederick

kfrederick

    Soyuz

  • *****
  • Posts: 3996
  • Joined: 01 Feb 2008

Posted 04 July 2014 - 11:23 AM

Correcting a 20 f 3.5 is a piece of cake from a spherical mirror that is how it is done . That is how I did this 36 inch f 3.5 the dark spots are fingerprints

Great! That's 38 waves of correction. How long did it take you to do that?

Not sure on the total but maybe 25 hrs here is my you tube of some of the correcting https://www.youtube....h?v=5dx3bncp8MM I used different sized laps I am pushing the arm by hand

#15 kfrederick

kfrederick

    Soyuz

  • *****
  • Posts: 3996
  • Joined: 01 Feb 2008

Posted 04 July 2014 - 11:27 AM

https://www.youtube....h?v=SnV0HdC4mHQ here is me doing a 24 inch f4 by hand

#16 Mark Harry

Mark Harry

    Cosmos

  • *****
  • Posts: 8121
  • Joined: 05 Sep 2005

Posted 05 July 2014 - 04:50 AM

Precisely. There's a BIG misconception that shallow, long focus mirrors are a cinch.... only if the maker knows exactly what he's doing!

#17 Pinbout

Pinbout

    ISS

  • *****
  • Posts: 22761
  • Joined: 22 Feb 2010

Posted 05 July 2014 - 08:16 AM

Danny, a 20" f/3.5 paraboloid and a sphere differ by 21 waves. Good luck parabolizing (figuring) that!



how many wave do you think this 8inf3.5 is...

Posted Image

took 4-5hours to get it close to spherical

Posted Image

#18 MKV

MKV

    Cosmos

  • *****
  • Posts: 7989
  • Joined: 20 Jan 2011

Posted 05 July 2014 - 09:50 AM

how many wave do you think this 8inf3.5 is...took 4-5 hours to get it close to spherical


The first one is not even a figure of revolution, so the question is pointless.
Mladen

#19 Pinbout

Pinbout

    ISS

  • *****
  • Posts: 22761
  • Joined: 22 Feb 2010

Posted 05 July 2014 - 03:37 PM

how many wave do you think this 8inf3.5 is...took 4-5 hours to get it close to spherical


The first one is not even a figure of revolution, so the question is pointless.
Mladen


Oh contraire Mon frere, it nulled with high zones. It is a figure of revolution. :p but since its on an AC they appear twice as high.

#20 MKV

MKV

    Cosmos

  • *****
  • Posts: 7989
  • Joined: 20 Jan 2011

Posted 06 July 2014 - 12:29 AM

Oh contraire Mon frere, it nulled with high zones. It is a figure of revolution.

If it's nulled (?), it would be a paraboloid with major surface roughness (macro ripple) of unknown amplitude (hello, it's a Ronchigram, not an inteferogram!), so how do you suggest you quantify this error in terms of waves?

And how does this relate to the difference of an f/3.5 paraboloid and a sphere, figure-wise? Apples and oranges, Danny.

Mladen

#21 Pinbout

Pinbout

    ISS

  • *****
  • Posts: 22761
  • Joined: 22 Feb 2010

Posted 06 July 2014 - 12:48 AM

If it's nulled (?)



I use that as a loose term cause it all greyed out at focus except you see the very high peaks.

how it relates is its a lot of glass that has to be removed with the polisher and cerox.

Its not a problem, its just time as in parabolizing your 20"f3.5.

but I tend to get figures off my MOM that are more corrected to -1 then 0 anyway.

Posted Image

#22 MKV

MKV

    Cosmos

  • *****
  • Posts: 7989
  • Joined: 20 Jan 2011

Posted 06 July 2014 - 01:31 AM

If it's nulled (?)



I use that as a loose term cause it all greyed out at focus except you see the very high peaks.

how it relates is its a lot of glass that has to be removed with the polisher and cerox.

Its not a problem, its just time as in parabolizing your 20"f3.5.

but I tend to get figures off my MOM that are more corrected to -1 then 0 anyway

Your image shows a rather gross surface roughness, but nowhere near the 9 waves (18 fringes) needed to bring an 8" f/3 sphere to a parabooid by deepeni

Mladen

#23 kfrederick

kfrederick

    Soyuz

  • *****
  • Posts: 3996
  • Joined: 01 Feb 2008

Posted 06 July 2014 - 05:37 AM

The reason to try to have a under corrected mirror at the end of primary polish is to have a good edge .Then shorten the RC of the whole mirror to match it .If you polish the center the RC shortens if you polish the area out from the .707 it gets longer .So three ways to correct a mirror . Polish the outer area increasing the final RC to match the center .Polish the edge and center and keep the same RC . Or call the edge good and polish the rest down and shorten final RC .The way most people who correct mirrors is to get a good edge and then deepen the whole mirror shorten the final RC .Hope that helps


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