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

8 inch, F/6 - Correction and Figuring

ATM mirror making
  • Please log in to reply
125 replies to this topic

#26 dogbiscuit

dogbiscuit

    Apollo

  • *****
  • Posts: 1,238
  • Joined: 18 Oct 2018
  • Loc: Barksdale

Posted 17 September 2020 - 12:32 PM

Was just now typing to ask for that. 

Looking pretty good.  The edge is better and zones might (probably will) go away with a little more work.


Astrokhan8post25Foucault.jpg

 

The zones were there before this work session, not as noticable either because of rough surface or turbulence.

The edge is better, most important

The parabolizing strokes will probably rub the zones out soon.

 

Send a picture of the lap.  I'll look to see if there is something about it that might cause zones.

I think it is just left from COC strokes but sometimes there is something in the lap shape.

 

The edge is better!  That's what is important now.  lol.gif   You make things look easy. 

 


  • Astrokhan likes this

#27 dogbiscuit

dogbiscuit

    Apollo

  • *****
  • Posts: 1,238
  • Joined: 18 Oct 2018
  • Loc: Barksdale

Posted 17 September 2020 - 12:44 PM

Do each of these in order 1 time around, then do them one more time and test

OneFourthWxThreeFourths.png

 

Parabolize4fifthsD7stroke.png

 

Don't do too much.  If things don't go right don't want to get far off track.

 

The big W should begin to remove the zones.


  • Astrokhan likes this

#28 Astrokhan

Astrokhan

    Vostok 1

  • -----
  • topic starter
  • Posts: 176
  • Joined: 01 Apr 2020
  • Loc: India

Posted 17 September 2020 - 01:32 PM

dogbiscuit, here is the current lap condition

 

IMG20200917235523-01.jpeg



#29 dogbiscuit

dogbiscuit

    Apollo

  • *****
  • Posts: 1,238
  • Joined: 18 Oct 2018
  • Loc: Barksdale

Posted 17 September 2020 - 02:53 PM

Looks pretty good.

 

Nit picking...

 

At the top of the lap, that channel is not needed at the edge.

on the right side, that one isn't needed either.  You could just not trim it and let it fill in.

 

See the central square is pretty small.  I would not trim it smaller than it is now. 

When trimming is needed trim the other side of the channel.

Over time as you work you can trim the sides of the channels that will work toward equal size facets and make them more square.

Not all at once, just each time you trim choose which side of the channel will work to make the lap better.

Astrokhan8lapPost28.jpg

 

 



#30 Astrokhan

Astrokhan

    Vostok 1

  • -----
  • topic starter
  • Posts: 176
  • Joined: 01 Apr 2020
  • Loc: India

Posted 18 September 2020 - 01:09 PM

Dogbiscuit, I worked further as suggested in #27. It seems to started getting parabolic but center hill is not going away. Am I missing lap contact at center? wanted to let you know that lap was without the mirror over it for several days before I started figuring and I had to do few long cold press to get the contact back. Though all squares looks in contact now but I am just guessing the possibility for center one. Let me know your views. Should I do a warm press before next session?

 

Inside RoC

IMG_4302cr_in.png

 

Outside RoC

IMG_4300cr_out.png



#31 Astrokhan

Astrokhan

    Vostok 1

  • -----
  • topic starter
  • Posts: 176
  • Joined: 01 Apr 2020
  • Loc: India

Posted 18 September 2020 - 01:17 PM

Here are two better KE images. The white splash between center and 5 o'clock is stains from dried water 

 

IMG_4334cr.png

IMG_4321cr.png



#32 Don H

Don H

    Apollo

  • -----
  • Posts: 1,245
  • Joined: 28 Dec 2006
  • Loc: Near Tucson, AZ

Posted 18 September 2020 - 02:41 PM

Just asking others to double check for you... Is the center square on the lap too close to the center? I always though the lap had to have squares well off center to avoid creating zones.

 

The good news is the edge is looking better, based upon the diffraction ring appearance...


  • Astrokhan likes this

#33 dogbiscuit

dogbiscuit

    Apollo

  • *****
  • Posts: 1,238
  • Joined: 18 Oct 2018
  • Loc: Barksdale

Posted 18 September 2020 - 07:24 PM

The central square is smaller than any on the lap.  That could have something to do with the hill.  And as Don H mentions it centered on the lap.

Even with all that the 4/5 D stroke the hill should have gotten smaller or gone away.

The narrow W  added some correction but not all the way to the edge, there is still some turn up in the outer zone.

The reason equal size squares on the lap are desired is because larger squares resist being pressed down during work more than small squares and they have more surface area to work the mirror. Pressure is higher and area more, more wear on the mirror.  The squares are far from uniform size.


I think what has happened is the narrow W that was to put correction in the outer zone, because it was near COC allowed the differences of facet size to generate the zones, making those zones faster than the big W can remove them.

I recommend you redo the lap.  I think it will cause trouble.  You might be able to force the mirror into shape using strokes to concentrate work on the high zones, or pressing parts of the lap out of contact (stars or rings), or accented pressure (usually TOT).  But that rarely results in as smooth a mirror as can be had if the lap is in good shape.


If you want to try to force the curve to better shape, here are some possible methods.

 

 

These large overhang strokes for working the center should work fast. I'd only do 1 turn of the mirror and see what happens. Would like the hill to come down smoothly, but if there is a little roughness there is time to smooth it as normal parabolizng strokes are used to get to the parabola.

It's the far off center part that works the mirror's center.  Take that sideways stoke at the ends of the W out close to the edge of the lap.

AstrokhanThirdx5sixths-2.png

 

This one leaves out all of the strokes except the far off center ones on one side. Red W is from 2/3 D out to the edge and magenta is from edge back into 2/3 D

astrokhanRemoveHill.png


For the narrow turn up at the edge, this accented pressure stroke will shorten the ROC of the zone just inside the turned up edge zone, so turn up is relatively less.  Pressure with right hand, left hand no pressure only helps guide the strokes. For the 8" about 2 or 3 pounds should work.  Don't do to much of this one either until you see how accented pressure works for you. The mirror is offset 1/2" to the right to starting position for the stroke.  The W is stroked to the right across a zone about 1" wide.  In the illustrations the yellow edge of the lap traces the stroking zigzag on the mirror surface.

astrokhanRemoveTueZone.jpg

 

Another thing might be to do a long narrow W like the one before except only 1/8 D wide. This could add correction in the outer 1/2" But it could also increase the zoning.

 

A better lap would be best but maybe you would like to try one or two of these first. 

You already know what sort of lap I like. smile.gif


  • Astrokhan likes this

#34 dogbiscuit

dogbiscuit

    Apollo

  • *****
  • Posts: 1,238
  • Joined: 18 Oct 2018
  • Loc: Barksdale

Posted 18 September 2020 - 08:07 PM

The accented pressure stroke is to push the zone between white lines down.

If you do this a second turn of the mirror move the stroke (both borders) about 3/8" toward center,

astrokhan8FoucaultPost31.jpg


  • Astrokhan likes this

#35 dogbiscuit

dogbiscuit

    Apollo

  • *****
  • Posts: 1,238
  • Joined: 18 Oct 2018
  • Loc: Barksdale

Posted 19 September 2020 - 12:49 AM

TOT version of accented pressure stroke.

Lap is offset to left and pressure applied to the right right edge of lap.

The right edge is stoked in a W pattern across the chosen zone.

For your current figure, If I did a second trip around the mirror I would shift the worked zone 3/8" toward the mirror's center.

 

astrokhanRemoveTueZoneTOT.jpg

 

 

 

 


  • Astrokhan likes this

#36 dogbiscuit

dogbiscuit

    Apollo

  • *****
  • Posts: 1,238
  • Joined: 18 Oct 2018
  • Loc: Barksdale

Posted 19 September 2020 - 01:38 AM

Foucault and Ronchi profile comparison.

astrokhan8profile31.jpg

 

 


  • Astrokhan likes this

#37 Astrokhan

Astrokhan

    Vostok 1

  • -----
  • topic starter
  • Posts: 176
  • Joined: 01 Apr 2020
  • Loc: India

Posted 19 September 2020 - 03:15 AM

Foucault and Ronchi profile comparison.

attachicon.gifastrokhan8profile31.jpg

dogbiscuit, you are indeed an expert and very sharp eyed. I am fortunate to be guided by you :)

 

I will be working further on your suggestion and post the results.



#38 Astrokhan

Astrokhan

    Vostok 1

  • -----
  • topic starter
  • Posts: 176
  • Joined: 01 Apr 2020
  • Loc: India

Posted 19 September 2020 - 03:51 PM

dogbiscuit, I decided to do the warm press and trim the lap. After that I did narrow 1/4D, followed by 4/5D and 3/4D as in #23 to smooth the surface. Now central hill is gone down and I think after some more work it will go away completely. See the pics below and advise for the next step.

 

I also made a video of KE and posted on youtube. please check and let me know if things are going well.

 

https://youtu.be/LOgmGyTroUc

 

Inside RoC

IMG_4341cr.png

 

Outside RoC

IMG_4351cr.png

 

KE

IMG_4354cr.png

 

 



#39 dogbiscuit

dogbiscuit

    Apollo

  • *****
  • Posts: 1,238
  • Joined: 18 Oct 2018
  • Loc: Barksdale

Posted 19 September 2020 - 07:11 PM

Your edge is good but the zones are bad. 

What does the lap look like now?

 

The high zone at the 70% zone is thing to work on.



#40 Astrokhan

Astrokhan

    Vostok 1

  • -----
  • topic starter
  • Posts: 176
  • Joined: 01 Apr 2020
  • Loc: India

Posted 20 September 2020 - 01:18 AM

Your edge is good but the zones are bad. 

What does the lap look like now?

 

The high zone at the 70% zone is thing to work on.

Here is the lap condition now. Tried to keep the channels as straight as possible but you can see the variation. The small edge facets you identified earlier are trimmed down from the pitch level. 

 

IMG20200920113859-01.jpeg



#41 dogbiscuit

dogbiscuit

    Apollo

  • *****
  • Posts: 1,238
  • Joined: 18 Oct 2018
  • Loc: Barksdale

Posted 20 September 2020 - 10:06 AM

It's looking better.

 

Compare to lap pattern B

astrokhan8lap40B.png

 

The narrow edge tile at center top and edge tile at center right are for me never less than 1/3 width of a full tile.

The red cross hair shows center of the lap circle.

The black x is the center of the most central facet of the B grid.

The most central tile of your lap is nearly centered on the lap circle.

 

I think it is that the center of your lap is very near the center of the central pitch square that is causing the zones.

The mirror is not far from a sphere. The high and low zones aren't all that high and low.

 

I don't know how much trouble you will have making a smooth parabola.  I guess you can try a few sessions to see if you can get the zones out.  If you don't make much progress in a few sessions, I'd say press until the channels fill in or make a new lap, channel again taking care to lay it out near perfect.

Note that the 37 mm shown in the lap grid drawing is between the center lines of the channels. The width of the pitch squares will be 37 mm minus 1 channel width so if your channels are 6 mm wide the pitch squares will be 31 mm.


  • Astrokhan likes this

#42 Astrokhan

Astrokhan

    Vostok 1

  • -----
  • topic starter
  • Posts: 176
  • Joined: 01 Apr 2020
  • Loc: India

Posted 21 September 2020 - 02:51 PM

dogbiscuit, I worked few more sessions and here is the current condition. Center look flatten now. Not sure about the edge but looks ok with slight turn up. Should I continue few more sessions? Let me know your views.

 

KE Video - https://youtu.be/jhThbXbi2o4

 

Inside RoC

 

IMG_4357cr.png

 

Outside RoC

 

IMG_4362cr.png

 

KE 

 

IMG_4368cr.png



#43 dogbiscuit

dogbiscuit

    Apollo

  • *****
  • Posts: 1,238
  • Joined: 18 Oct 2018
  • Loc: Barksdale

Posted 22 September 2020 - 07:13 AM

Still kind of zoney.

You should be taking knife readings before adding too much correction.

Fixing those zones could add correction and could end up over corrected.

W 2/3D stroke length 1" wide W 2 times around and no more!

Too much will be a bad thing.  It might just be a bad thing.  So don't do much of it.

Test.

Knife readings.

Pictures, mirror and lap

How thick is the lap?


Edited by dogbiscuit, 22 September 2020 - 07:16 AM.

  • Astrokhan likes this

#44 dogbiscuit

dogbiscuit

    Apollo

  • *****
  • Posts: 1,238
  • Joined: 18 Oct 2018
  • Loc: Barksdale

Posted 22 September 2020 - 07:18 AM

The lap is the figuring secret sauce.


  • Astrokhan likes this

#45 Astrokhan

Astrokhan

    Vostok 1

  • -----
  • topic starter
  • Posts: 176
  • Joined: 01 Apr 2020
  • Loc: India

Posted 22 September 2020 - 08:15 AM

The lap is the figuring secret sauce.

very true!



#46 Astrokhan

Astrokhan

    Vostok 1

  • -----
  • topic starter
  • Posts: 176
  • Joined: 01 Apr 2020
  • Loc: India

Posted 22 September 2020 - 08:18 AM

Still kind of zoney.

You should be taking knife readings before adding too much correction.

Fixing those zones could add correction and could end up over corrected.

W 2/3D stroke length 1" wide W 2 times around and no more!

Too much will be a bad thing.  It might just be a bad thing.  So don't do much of it.

Test.

Knife readings.

Pictures, mirror and lap

How thick is the lap?

Ok. I will work as suggested and post the results. I have not yet taken any knife readings and just doing visual observation of figure. Not yet made the couder mask. should I make one and start taking the readings?

Lap is around 5-7 mm thick at the edge.



#47 dogbiscuit

dogbiscuit

    Apollo

  • *****
  • Posts: 1,238
  • Joined: 18 Oct 2018
  • Loc: Barksdale

Posted 22 September 2020 - 12:36 PM

I don't like the the relatively recent trend to do all or most of figuring only by Ronchi test.

It's better to take knife readings through all figuring.

The numbers will give you a better knowledge of the state and progression of correction.

If I were only going to do one test Foucault or Ronchi it would be Foucault.

 

Relative to the entire mirror there is a lot of correction in the zone between 55% and 85% radii. 

how much correction is it?  Hard to tell with Ronchi.

 

If you keep correction even across the entire mirror it's easier to judge degrees of correction.

When there are zones it is difficult.  You have zones.

 

I think you zones are severe enough that you should work toward a sphere and quite likely you will need a better lap to get there.  Besides the bad shape, your lap is thinner than what I like.  For me I'd like about 10 mm thickness at start of parabolizing and if the figure isn't completed when the lap is around 6 mm the lap is finished and a new lap is needed.

But that's me.  Maybe it will work different for you.



#48 Steve Dodds

Steve Dodds

    Owner - Nova Optical

  • *****
  • Vendors
  • Posts: 607
  • Joined: 02 Apr 2007
  • Loc: Utah

Posted 22 September 2020 - 12:54 PM

I don't like the the relatively recent trend to do all or most of figuring only by Ronchi test.

It's better to take knife readings through all figuring.

The numbers will give you a better knowledge of the state and progression of correction.

If I were only going to do one test Foucault or Ronchi it would be Foucault.

 

Relative to the entire mirror there is a lot of correction in the zone between 55% and 85% radii. 

how much correction is it?  Hard to tell with Ronchi.

 

If you keep correction even across the entire mirror it's easier to judge degrees of correction.

When there are zones it is difficult.  You have zones.

 

I think you zones are severe enough that you should work toward a sphere and quite likely you will need a better lap to get there.  Besides the bad shape, your lap is thinner than what I like.  For me I'd like about 10 mm thickness at start of parabolizing and if the figure isn't completed when the lap is around 6 mm the lap is finished and a new lap is needed.

But that's me.  Maybe it will work different for you.

I couldn't disagree more.  With the focault test it's 15-20 minutes to get a reliable set of readings, with the Ronchi it's 3 seconds.  With the Ronchi it's fairly easy to get within 1/2 wave, then and only then switch to a knife edge.


  • LarsMalmgren and sopticals like this

#49 dogbiscuit

dogbiscuit

    Apollo

  • *****
  • Posts: 1,238
  • Joined: 18 Oct 2018
  • Loc: Barksdale

Posted 22 September 2020 - 01:30 PM

I couldn't disagree more.  With the focault test it's 15-20 minutes to get a reliable set of readings, with the Ronchi it's 3 seconds.  With the Ronchi it's fairly easy to get within 1/2 wave, then and only then switch to a knife edge.

I'm not in so much of a hurry that 15 or 20 minutes.

Might be able to see Ronchi bands in 3 seconds but it take longer to interpret.

For some mirrors I could get within 1/2 wave with only Foucault without the mask as quickly as with Ronchi.

Without the mask ball park knife readings can be taken between a few zones, and there are things about how the null zone moves, abnormal speeding up or slowing down as the knife is moved between center null and outer zone null that show kinks in the curve.

Knife readings with a mask are not so difficult and can give better understanding of the current state of the mirror.


  • Etoile likes this

#50 dogbiscuit

dogbiscuit

    Apollo

  • *****
  • Posts: 1,238
  • Joined: 18 Oct 2018
  • Loc: Barksdale

Posted 23 September 2020 - 01:49 AM

Not knowing the amount of correction between 55 and 85%, if you just wanted to deepen the flat center to match, one or two turns of this MOT stroke should do it.  Large overhang stroke like this can work the center real fast, I'd do one turn at a time and test. If that doesn't over correct the mirror and the center blends well with the 55-85 zone might be close to a parabola. Then there would only be the raised edge.    I put the Foucault image in to show you where the edge of the lap wears the mirror's center.

 

Astrokhan42CenterCorrection.png

 

Can't say you should do this.  There is a knack to getting smooth results with this sort of large overhang stroke.


  • Astrokhan 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





Also tagged with one or more of these keywords: ATM, mirror making



Cloudy Nights LLC
Cloudy Nights Sponsor: Astronomics