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

Correcting a 16" flat to correct a 16" paraboloid

  • Please log in to reply
82 replies to this topic

#1 ccaissie

ccaissie

    Surveyor 1

  • *****
  • topic starter
  • Posts: 1767
  • Joined: 13 Sep 2010
  • Loc: Whitefield, Maine

Posted 31 October 2019 - 04:50 AM

Our club is interested in correcting a 16" f/4.5 Meade primary mirror that suffers from astigmatism, and a "less than optimal" figure (see FigureXP analysis).  I had refigured a similar mirror for our observatory, which performs very well, using Foucault.  We swap out the good mirror into our dob for travel to star parties...most recently to STARS OVER KATAHDIN...man, it's dark out there, one of THE DARKEST sites east of the Mississippi, and we're fighting to keep it that way.

 

In this case, there is astigmatism showing up in the star test, and it is certainly on the primary.  I refigured the secondary to VERY flat, and the "good" mirror #1 shows zero stig.

 

I have a full thickness 16" dia. perforated Pyrex "flat" that has some power and some zonal issues, shown in the pics.

 

To "catch two fish on the same line" (rather than killing birds with a rock), I thought I'd do some touchup on the flat and then DPAC the mirror.

 

The quick test on half of the flat shows 2+ waves (4-5 fringes) CONCAVE, using our 10" Henry Paul 1/10th wave flat. I assume it's more like 6-8 fringes over 16".  I'm assuming that the power of the flat is negligible, but the irregularity, lack of smoothness is not.  The flat must have been used for some collimation tests..."flat" on one side, f/7 sphere on the back.  

 

In the community's expert opinion, is this flat suitable for retouching, and how would I best go about it?  I'm concerned about the irregularity beginning at the 5-6" radius zone.  I assume it is useless as-is for correcting a paraboloid DPAC.  The stig issue will be handled by polishing the mirror with a good sized lap and using pinhole Foucault technique.

 

C

Attached Thumbnails

  • IMG_2611_1.JPG
  • IMG_2613_1.JPG
  • IMG_2615_1.JPG
  • 16f4_5 No 2Surface.gif

Edited by ccaissie, 31 October 2019 - 04:59 AM.


#2 Mark Harry

Mark Harry

    Cosmos

  • *****
  • Posts: 8280
  • Joined: 05 Sep 2005
  • Loc: Northeast USA

Posted 31 October 2019 - 07:10 AM

To me, it looks like 3.5-4 fringes of power, and if it's to be used for AC the fringes = waves error, likely a non issue.
Second pic, looks like it was taken too close without the return image being collimated enough???
The 3rd looks better when you were further back.
If the first image "central ring" can be viewed and assessed for nice roundness, that would show stig if it was deformed in any way. If you force that central ring toward the edge, it may distort, giving you a false result, just saying.


  • MKV, Eikonal and Gleb1964 like this

#3 Pinbout

Pinbout

    ISS

  • *****
  • Posts: 22989
  • Joined: 22 Feb 2010

Posted 31 October 2019 - 08:43 AM

It can have power. 

 

the smoother the better, you are showing a shallow kink.

 

did you set it up for dpac yet?

 

Should be fine especially for seeing overall correction



#4 ccaissie

ccaissie

    Surveyor 1

  • *****
  • topic starter
  • Posts: 1767
  • Joined: 13 Sep 2010
  • Loc: Whitefield, Maine

Posted 31 October 2019 - 10:30 AM

Not setup in DPAC.  Pic 1 shows the 10" flat contacting and showing the general concavity...Not too rough, but not circular, shows the power.

 

Pic 2 tipped to show 'straight' fringes to reveal zones, and they are there...that's what I'm concerned about...how much the deviation from a true curve will throw a false reading for that zone when DPAC.  

 

Certainly not "collimated". I was hovering about 4' over the rig...variation in scale is from zooming a bit.

 

I think I'll ponder the 'kink' in Pic 2, and see if I can get a sense of how far from a smooth curve it is.  

Question...if it looks like a smooth arc can I assume it will do for some close DPAC correcting?  I remember a hot topic earlier about the power of a flat not being very sensitive to whether the power was spherical or aspherical....don't want to touch that one...

In this case it looks like it's generally smooth concave until it hits 5-6" radius, kinks, then is pretty flat...fringes go straight to the edge, which is in good shape.   

 

In any case, looks like some general smoothing with some softer pitch, say a 10" lap,  relying on stroke to get more action over the 5-7" radius...    or would y'all begin with a 6" over the kink?  "Deepen" the 6-7" zone to blend it in with the prevailing concavity?

 

I did have some difficulty on smoothness when I was correcting a 10" flat last Winter..lap was on the hard side.  Was "flat" but not smooth.

 

I will deal with the parallax-distance setup issue to get better purity in the future tests.  This was a quickie.


Edited by ccaissie, 31 October 2019 - 10:40 AM.


#5 Mike Spooner

Mike Spooner

    Vendor (mirrors)

  • *****
  • Vendors
  • Posts: 607
  • Joined: 06 Aug 2010

Posted 31 October 2019 - 11:11 AM

If you have a known good scope or mirror, set it up and run a test. This can be instructive.


  • PrestonE, Mark Harry, John Lightholder and 2 others like this

#6 PeteDCard81

PeteDCard81

    Explorer 1

  • -----
  • Posts: 52
  • Joined: 30 Nov 2017
  • Loc: Twin Cities, Minnesota

Posted 31 October 2019 - 11:16 AM

If it was me trying to better the flat I would use a 20-inch pitch lap, with the appropriate softness, polishing with the flat on top. This will make it more convex. You could then use the other 16-inch mirror with the better figure to test the flat in DPAC. If the flat became convex just hang it over the edge to deepen it.

 

To get a true picture of the astigmatism on the mirror that you want to correct you can use a Bath Interferometer and DFTFringe.

 

 

Mark


Edited by PeteDCard81, 31 October 2019 - 06:33 PM.


#7 ccaissie

ccaissie

    Surveyor 1

  • *****
  • topic starter
  • Posts: 1767
  • Joined: 13 Sep 2010
  • Loc: Whitefield, Maine

Posted 31 October 2019 - 04:44 PM

If you have a known good scope or mirror, set it up and run a test. This can be instructive.

Yes, I suppose the errors on the flat would show against what I already know about mirror #1, which is good...HOW GOOD?   Well there's the rub, eh?  Are the errors on the flat or the mirror?

 

Re: 20" lap....well, I'm not trying to make it flat, just smooth for DPAC...A 20" won't fit on my spindle anyways...

 

I could set up the Ritchie-Common test..I made an 8" f/15 sphere that I know is beyond reproach....


Edited by ccaissie, 31 October 2019 - 04:47 PM.


#8 MKV

MKV

    Cosmos

  • *****
  • Posts: 8164
  • Joined: 20 Jan 2011
  • Loc: Florida, USA

Posted 31 October 2019 - 04:58 PM

...that's what I'm concerned about...how much the deviation from a true curve will throw a false reading for that zone when DPAC.

Mark Harry gave you the answer -- get a bulls eye fringe pattern with as few fringes as possible and observe how smooth and circular the cenetral fringe is. If it's circular the figure is good.

 

As for how many fringes of power you can have, the tolerance was established back in 1938 by C. R. Burch, who worked out a 1/20 wave tolerance in terms of fringes of power for a given focal ratio. It's a lot more "generous" than most people think.

 

http://adsabs.harvar...MNRAS..98..670B

 

The short of it basically says the maximum acceptable number of fringes = 3.6*(F#)2, where F# is your mirror's focal ratio.  Thus, for an f/2 mirror the allowed number of fringes in an autocollimation flat should not be more than 3.6*4 = 14, which equals 7 waves.

 

For your f/4.5, the maximum number of fringes allowed = 3.6*(4.5)2 = 74 or 37 waves. Prudence, however, tells us to shoot for as few fringes as possible.


  • Oregon-raybender likes this

#9 ccaissie

ccaissie

    Surveyor 1

  • *****
  • topic starter
  • Posts: 1767
  • Joined: 13 Sep 2010
  • Loc: Whitefield, Maine

Posted 31 October 2019 - 05:06 PM

Mark Harry gave you the answer -- get a bulls eye fringe pattern with as few fringes as possible and observe how smooth and circular the cenetral fringe is. If it's circular the figure is good.

 

As for how many fringes of power you can have, the tolerance was established back in 1938 by C. R. Burch, who worked out a 1/20 wave tolerance in terms of fringes of power for a given focal ratio. It's a lot more "generous" than most people think.

 

http://adsabs.harvar...MNRAS..98..670B

 

The short of it basically says the maximum acceptable number of fringes = 3.6*(F#)2, where F# is your mirror's focal ratio.  Thus, for an f/2 mirror the allowed number of fringes in an autocollimation flat should not be more than 3.6*4 = 14, which equals 7 waves.

 

For your f/4.5, the maximum number of fringes allowed = 3.6*(4.5)2 = 74 or 37 waves. Prudence, however, tells us to shoot for as few fringes as possible.

Yes, the circularity will tell me if it's warped or really rough, but it won't tell me if there's a concentric zone, which the performed contact interference test DOES show....  unless you think the departure shown in the bands (the almost straight ones, pic 2) are straight enough to ignore that kink... I'd be happy to be able to ignore that kink and go on to DPAC.  I dig the math that shows my 5-8 fringes is laughable...



#10 BGRE

BGRE

    Vanguard

  • *****
  • Posts: 2479
  • Joined: 21 Mar 2016
  • Loc: New Zealand

Posted 31 October 2019 - 05:10 PM

Why not a Rayleigh water test? If a suitable collimator isn't available its possible to calculate the wavefront for a perfect flat from the test geometry and subtract this from the measured results:

 

 C3 =  (1/8)*N*t*(1 - 2* (t/L))*(D/L)^2 + (3/128)*N*t*(1-4*(t/L))*(D/L)^4

C8 = (1/128)*N*t*(1 - 4*(t/L))*(D/L)^4

L is the distance from the camera lens entrance pupil to the test surface

t is the thickness of the water film

D is the test surface diameter

N is the refractive index of water.



#11 Pinbout

Pinbout

    ISS

  • *****
  • Posts: 22989
  • Joined: 22 Feb 2010

Posted 31 October 2019 - 05:56 PM

 

Yes, I suppose the errors on the flat would show against what I already know about mirror #1, which is good...HOW GOOD?   Well there's the rub, eh?  Are the errors on the flat or the mirror?

A mirror that is smooth, no kinks - the wavefront isn’t important if you want to see if the kink on the flat will show on a smooth parabolic that’s about 1/2~

 

You can wipe the 1 line ronchi left to right, flip it inside outside focus...



#12 Vla

Vla

    Viking 1

  • *****
  • Posts: 611
  • Joined: 08 Mar 2014

Posted 31 October 2019 - 06:01 PM

I would call 0.8 wave p-v of spherical aberration "much less than optimal" figure. My guess is that Figure XP is set to show how to get to a paraboloid with the least amount of glass removed. In this case, it is reverse parabolizing, by flattening the 0.7 zone, which would mean that the figure is a hyperboloid with turned up edge (about 94% zone). The wavefront error indicates -1.2 conic.

 

As for the astigmatism, a random thought: would it be possible to fix it with astigmatic diagonal?

Attached Thumbnails

  • HPS.png


#13 Ed Jones

Ed Jones

    Soyuz

  • *****
  • Posts: 3866
  • Joined: 06 Apr 2004
  • Loc: Sin-sin-atti

Posted 31 October 2019 - 06:43 PM

Pic #1 is probably too small a flat.  #2 definitely shows some irregularity and #3 shows some asymmetry.  In contact testing the fringes should always be perfect circles and irregularity shows as deviations from perfect circles.  What are the small circles?  I hope not shims.

 

Hows the secondary?  It's more likely the problem.


Edited by Ed Jones, 31 October 2019 - 06:44 PM.


#14 ccaissie

ccaissie

    Surveyor 1

  • *****
  • topic starter
  • Posts: 1767
  • Joined: 13 Sep 2010
  • Loc: Whitefield, Maine

Posted 31 October 2019 - 07:00 PM

OK, here's a COC 100lpi Ronchi of the bad 16" mirror I want to refigure.  It was so bad I really didn't detect astigmatism until I refigured the secondary and again star tested it.

 

Again just looking at it says that the kink at the ~70% zone has to go and should be smooth to the eye.  That's where Ronchi is so great, reading the smoothness.  The DPAC null is the final test.  I recently had a few mirrors tested at Stellafane..the DPAC Nulls were excellent.  But this is a bigger boy, and needs a 16" flat.

 

Ed:

The small circles...one circle is the 1.3" central perforation of the 16 flat, showing in all photos.  The other "circle" is the ground in "moon" on the back edge of the 10" flat...Henry Paul included it on that one...good idea...helps in identifying at a glance which surface is the flat.

These discs are contacted, no spacers, and they have been associated for about a month. definitely equalized. I've checked on them often to see if temp change has anything to do with figure...nope...same- same.

 

I refigured the secondary, which was about 1/4 wave concave to a much better figure...it's oversized at 4".  I think astigmatizing it to correct for the primary is "against all optical instinct".  -quote attributed to Conrady

 

Pic 2 is the one that clues me in to irregularity...if the arcs were even, I'd not hesitate to use it as-is.  In that pic, the straightedge goes from the central perforation to the edge of the 16 flat.

Attached Thumbnails

  • post-142038-0-18210600-1518137020_thumb.jpg

Edited by ccaissie, 31 October 2019 - 07:16 PM.


#15 Ed Jones

Ed Jones

    Soyuz

  • *****
  • Posts: 3866
  • Joined: 06 Apr 2004
  • Loc: Sin-sin-atti

Posted 31 October 2019 - 07:54 PM

I don't see astigmatism in this Ronchigram but the Ronchi isn't a good test at all for astigmatism, it does show a bad kink.   If it's still coated do autocollimation off an oil flat.



#16 Vla

Vla

    Viking 1

  • *****
  • Posts: 611
  • Joined: 08 Mar 2014

Posted 31 October 2019 - 08:29 PM

I refigured the secondary, which was about 1/4 wave concave to a much better figure...it's oversized at 4".  I think astigmatizing it to correct for the primary is "against all optical instinct".  -quote attributed to Conrady

So is the Schmidt, Maksutov, Ritchey-Chretien... A 2800m radius (sphere, little over 1/3 wave deep) on the diagonal at 1.5m from the mirror would produce 1/2 wave p-v of astigmatism. Twice as deep, twice as much. The rest is just properly orienting the primary.



#17 MKV

MKV

    Cosmos

  • *****
  • Posts: 8164
  • Joined: 20 Jan 2011
  • Loc: Florida, USA

Posted 01 November 2019 - 12:35 AM

That's where Ronchi is so great, reading the smoothness.  The DPAC null is the final test.

Why not use AC from the start -- it's easier to interpret and it's double sensitive?



#18 Mark Harry

Mark Harry

    Cosmos

  • *****
  • Posts: 8280
  • Joined: 05 Sep 2005
  • Loc: Northeast USA

Posted 01 November 2019 - 02:39 AM

I still say the #2 pic may, or may NOT be an issue. The pic was made with camera awful close to the image, and might just be showing distortion.
(indicated by decent arcs showing in #3)

As for the 10" flat, it's only 1/2 as sensitive as a full size 16" would be. So that isn't a total loss either. Colin should have no trouble using the 10" reference.
*******
Regardless, the flat is smoother than that pukey looking "raunchy-kink" on the mirror itself. I'd start correcting that mess anyhow; heck it's got to improve!


  • MKV likes this

#19 Pinbout

Pinbout

    ISS

  • *****
  • Posts: 22989
  • Joined: 22 Feb 2010

Posted 01 November 2019 - 11:32 AM

 

OK, here's a COC 100lpi Ronchi of the bad 16" mirror I want to refigure

yep, you don't need a flat to correct this kink... fix the kink then test with the flat.

 

figure 8's / ovals, short front to back,  not wide left to right, full size tot... so not much hang over when the tot is hanging over... grin.gif


  • PrestonE likes this

#20 ccaissie

ccaissie

    Surveyor 1

  • *****
  • topic starter
  • Posts: 1767
  • Joined: 13 Sep 2010
  • Loc: Whitefield, Maine

Posted 01 November 2019 - 03:49 PM

I still say the #2 pic may, or may NOT be an issue.

 

Regardless, the flat is smoother than that pukey looking "raunchy-kink" on the mirror itself. I'd start correcting that mess anyhow; heck it's got to improve!

I will retouch that issue showing on the flat in pic 2.  Carefully.  It's like a 1/6 wave zonal error in a 16" 8 fringe Concave flat.

The measure of success in this venture is that I wind up with an excellent mirror AND a very usable flat.

 

Appreciate Danny's stroke recipe, especially full lap...that's for the 'stig' on that that ugly mirror...stand back, hold ma beer....



#21 DAVIDG

DAVIDG

    Cosmos

  • *****
  • Posts: 9081
  • Joined: 02 Dec 2004
  • Loc: Hockessin, De

Posted 01 November 2019 - 03:58 PM

Having fixed a couple of flats this is what I do. Make a 10" lap that is only about 1/4" deep, and with 1" square facets. Press it really well against your 10" Master. Now use a figure 8 stroke with the Flat of course on the bottom. Make sure  the figure 8 stroke cover the complete surface of the flat ie top to bottom right to left but with only a 1/2" or less over hang.

Test after 10 minutes and keep pressing the lap against the Master. Make sure the channels stay open to the same width across the lap so the facets stay the same size. Equal facet size equals equal wear and the surface will go flat.

 

              - Dave 


  • PrestonE and Pinbout like this

#22 ccaissie

ccaissie

    Surveyor 1

  • *****
  • topic starter
  • Posts: 1767
  • Joined: 13 Sep 2010
  • Loc: Whitefield, Maine

Posted 01 November 2019 - 04:41 PM

I don't see astigmatism in this Ronchigram but the Ronchi isn't a good test at all for astigmatism, it does show a bad kink.   If it's still coated do autocollimation off an oil flat.

Star test shows ...guessing...1/4 wave stig in a mirror that's almost 1 wave off...granted, refocus and central obstruction reduces the announced p-v error, but it's a bad mirror...



#23 ccaissie

ccaissie

    Surveyor 1

  • *****
  • topic starter
  • Posts: 1767
  • Joined: 13 Sep 2010
  • Loc: Whitefield, Maine

Posted 01 November 2019 - 04:45 PM

Having fixed a couple of flats this is what I do. Make a 10" lap that is only about 1/4" deep, and with 1" square facets. Press it really well against your 10" Master. Now use a figure 8 stroke with the Flat of course on the bottom. Make sure  the figure 8 stroke cover the complete surface of the flat ie top to bottom right to left but with only a 1/2" or less over hang.

Test after 10 minutes and keep pressing the lap against the Master. Make sure the channels stay open to the same width across the lap so the facets stay the same size. Equal facet size equals equal wear and the surface will go flat.

 

              - Dave 

Yes, I wouldn't want to even touch that edge...it's nice.  reluctant to try to flatten the disk...risk of flattening the edge and not getting the 2-4wave of concavity out, buggering the edge and making me unhappy...less is more for me here.



#24 Vla

Vla

    Viking 1

  • *****
  • Posts: 611
  • Joined: 08 Mar 2014

Posted 01 November 2019 - 04:47 PM

What figure XP shows is a conic error (spherical aberration) which can't be reduced by refocusing. What it shows is a hyperboloid (-1.2) with 0.4 wave turned up edge beginning at about 94% zone.



#25 Pinbout

Pinbout

    ISS

  • *****
  • Posts: 22989
  • Joined: 22 Feb 2010

Posted 01 November 2019 - 06:56 PM

Having fixed a couple of flats this is what I do. Make a 10" lap that is only about 1/4" deep, and with 1" square facets. Press it really well against your 10" Master. Now use a figure 8 stroke with the Flat of course on the bottom. Make sure  the figure 8 stroke cover the complete surface of the flat ie top to bottom right to left but with only a 1/2" or less over hang.

Test after 10 minutes and keep pressing the lap against the Master. Make sure the channels stay open to the same width across the lap so the facets stay the same size. Equal facet size equals equal wear and the surface will go flat.

 

              - Dave 

What if it’s already perforated?




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