Telescope Reviews: Refiguring 16" f/4.5 Meade

· Get a Cloudy Nights T-Shirt

· Submit a Review / Article

Click here if you are having trouble logging into the forums

Privacy Policy | Please read our Terms of Service | Signup and Troubleshooting FAQ | Problems? PM a Red or a Green Gu… uh, User

You are not logged in. [Login]

Entrance · Main Index · Search · New user · Who's Online
FAQ · Calendar

Equipment Discussions >> ATM, Optics and DIY Forum

Pages: 1 | 2 | 3 | 4 | (show all)

ccaissie
sage

Reged: 09/13/10

Loc: Whitefield, Maine

Refiguring 16" f/4.5 Meade
#5400743 - 09/03/12 09:30 AM Attachment (85 downloads)

As if I didn't have enough projects to keep me distracted...I've decided it's time to recoat the 16 in the club's observatory. Since the performance after my last refiguring was much improved, now I'm nitpicky, and want to give it a few additional swipes with a pitch lap.

When we assembled the scope as an Equatorial using a Meade 16 Dob, the mirror was very much overcorrected in the star test...no snap to the focus, etc.

At it's first recoat, I Foucault/pinstick tested it and found it was almost a wave overcorrected. Several figuring sessions later yielded what I thought was about 1/6 wave overcorrected. When we put it back in the scope, it seemed UNDERcorrected about 1/4 wave, and has been a fair performer.

But now that it's back on the bench, initial Foucault/pinstick says 1/4 wave overcorrected.

My question is: If Foucault is the better measure, then why does it exhibit star test undercorrection?

1. My foucault testing is suspect.
2. Cooldown, tube currents, etc. are subverting the star test...I get slightly different results on some evenings. Scope has a SOLID tube.
3. Some combination of higher spherical abberation has got me confused re: star test.

Anyway, I'm blogging this process here, and ask for feedback.

My next steps are to strip the coating, recheck everything and test, test, test. I'm considering the Ross, and also using a 12" flat to partially get some objective autocollimation readings. I also have a 16" "flat" that is actually about 10 waves concave, but who knows how smooth it is beneath all those fringes.
Our club has a 10" tenth wave flat made and signed by Henry Paul, but it's maybe not large enough to use for correcting the 16 flat, which is on a list of:

Too many projects...

Post Extras:

Print Post

Remind Me!

Notify Moderator

Shane LaPierre
super member

Reged: 04/17/09

Re: Refiguring 16" f/4.5 Meade new

new

[Re: ccaissie]
#5400777 - 09/03/12 09:49 AM

Too many projects, indeed

Trust the Foucault, but also trust the star test. Add a Ronchi test. I think that can help rule out zones. Also, try ronchi at the focus (on a star). It's hard to get good seeing, but over a period of a few nights I can usually get a good picture of the overall shape.

Good luck.

Post Extras:

Print Post

Remind Me!

Notify Moderator

MKV
Carpal Tunnel

Reged: 01/20/11

Re: Refiguring 16" f/4.5 Meade new

new

[Re: ccaissie]
#5400844 - 09/03/12 10:38 AM

First, do you have any pictures? Second, you have every reason not to trust the Foucault at that aperture and focal ratio. A 16 inch f/4.5 must be better than 99.8% parabola. That leaves you less than ± 0.2% of "wiggle" room. Foucault is pathetically undersampled and insensitive when it comes to large, fast mirrors.

Alternatives are, in order of preference, autcollimation, Ross null test, interferometry. All have their disadvantages for different reasons. Autcollimation requires a flat bigger than 16 inches. Good luck!!!

Ross requires very accurate lens-mirror separation (it also requires a good quality large pcx lens with at least 300 mm focal length).

Interferometry is usually not in the amateur domain unless you make a good Bath interferometer and meticulously calibrate it.

There is also a conjugate null test that is simialr to the Ross null, but can be done with a smaller lens. However, unlike the Ross null test for which there is a subroutine called RossNullXP, the exact positioning of the light source and the lens-to-knife-edge/Ronchi grating separation for the conjugate null is best done by raytracing.

Also, with the Ross/conjugate method, if the separations are not accurate, you will still get a null, but it won't be a paraboloid! (that's what happened to the Hubble Space Telescope's original mirror)

Calibration of all test equipment is a must. You have to know the internal reading error of your equipment. You may be surprised how much of that error is present - probably enough to send that ± 0.2% into the never-never land.

If you see hills and bumps, continue figuring the mirror until they are gone and you have a smooth surface, then figure it as well as you can with the traditional methods. Then send the mirror to a reputable professional company and have it tested.

Regards,

Mladen

Post Extras:

Print Post

Remind Me!

Notify Moderator

ccaissie
sage

Reged: 09/13/10

Loc: Whitefield, Maine

Re: Refiguring 16" f/4.5 Meade new

new

[Re: MKV]
#5400883 - 09/03/12 11:08 AM Attachment (74 downloads)

No pictures yet, other than initial setup. Some zoniness near the edge with Ronchi.

All helpful comments. I get into these optical thingies infrequently and it takes a bit to get back into the swing. My last project was a 12.0 f/3 paraboloid for an astrograph that tested better than 1/5, smooth and good edge. Autocollimation was the ticket there. 10" f/8 at 1/15 wave was next with Foucault, but I hear you...16" f/4.5 is another animal.

I've star tested an 18" Pegasus which shows me what a better mirror is, so I do have some guidelines. I've also star and bench tested other optics so there's some experience there.

Post Extras:

Print Post

Remind Me!

Notify Moderator

ccaissie
sage

Reged: 09/13/10

Loc: Whitefield, Maine

Re: Refiguring 16" f/4.5 Meade new

new

[Re: Shane LaPierre]
#5400905 - 09/03/12 11:26 AM

Have gotten good results at the eyepiece using a small Mars instead of a star.

Post Extras:

Print Post

Remind Me!

Notify Moderator

Benach
Pooh-Bah

Reged: 01/24/08

Re: Refiguring 16" f/4.5 Meade new

new

[Re: ccaissie]
#5400997 - 09/03/12 12:25 PM Attachment (67 downloads)

And this difference doesn't surprise you?

Post Extras:

Print Post

Remind Me!

Notify Moderator

DAVIDG
Post Laureate

Reged: 12/02/04

Loc: Hockessin, De

Re: Refiguring 16" f/4.5 Meade new

new

[Re: ccaissie]
#5401015 - 09/03/12 12:37 PM

I believe that the more test methods the better and most importantly all the results should agree, if not ones needs to figure out why they don't. Is it the tester, is it your technique, it is a wrong measurement? Also one needs to understand that each one of the readings has an error associated with it. Without knowing what that error is, you don't know what your confidenece level is in the final results. Is the surface you measured 1/8 wave with possible error of +/- 1/20 or is it 1/8 wave with an unknown error of +/- 1/4 wave ?
You can easily setup a Ross Null. The lens doesn't need to be perfect and it will allow you to view the complete surface which may show you problems missed with knife edge readings.
Also since you have a complete OTA, you can use a smaller diameter flat then the size of the mirror and do double pass autocollimation. The optics are reinstalled in the tube assembly and the smaller flat placed in front of the scope and off to one side. You'll be able to get a good look at the 70% zone and out to the edge which are the most critical parts of the mirror.

- Dave

Post Extras:

Print Post

Remind Me!

Notify Moderator

ccaissie
sage

Reged: 09/13/10

Loc: Whitefield, Maine

Re: Refiguring 16" f/4.5 Meade new

new

[Re: DAVIDG]
#5401024 - 09/03/12 12:45 PM

Quote:

Also since you have a complete OTA, you can use a smaller diameter flat then the size of the mirror and do double pass autocollimation. The optics are reinstalled in the tube assembly and the smaller flat placed in front of the scope and off to one side. You'll be able to get a good look at the 70% zone and out to the edge which are the most critical parts of the mirror.- Dave

Blinding Flash of the OBVIOUS!

That's exactly what I needed, a way to include autocollimation into the test array.

Post Extras:

Print Post

Remind Me!

Notify Moderator

alnitak
Vendor (Alnitak Astrosystems)

Reged: 10/12/05

Re: Refiguring 16" f/4.5 Meade new

new

[Re: ccaissie]
#5401362 - 09/03/12 04:03 PM

Colin, I thought mine was the last!

Post Extras:

Print Post

Remind Me!

Notify Moderator

MKV
Carpal Tunnel

Reged: 01/20/11

Re: Refiguring 16" f/4.5 Meade new

new

[Re: Benach]
#5401927 - 09/03/12 09:46 PM

Quote:

And this difference doesn't surprise you?

That makes it only 97.2% parabola, assuming, of course that the radius of curvature is 143 ± 0.0000 inches! Not to mention the zonal mask radii...cut by hand with a packing knife?

Regards,

Mladen

Edited by MKV (09/03/12 09:57 PM)

Post Extras:

Print Post

Remind Me!

Notify Moderator

mark cowan
Vendor (Veritas Optics)

Reged: 06/03/05

Loc: salem, OR

Re: Refiguring 16" f/4.5 Meade new

new

[Re: Shane LaPierre]
#5402206 - 09/04/12 12:50 AM

Quote:

Trust the Foucault, but also trust the star test.

Yes, this. Work on both of them until you get agreement. Otherwise, what are you doing? Putting changes on the glass based on an unreliable test. And then reading those changes with an unreliable test. First establish cause and effect.

Ronchi at focus isn't very useful unless you're looking at just a line or two, in which case the grating needs to be high quality. Ronchi on the bench can tell you quite a lot about smoothness and will help you discover zones, but not quantify them. Ross null is easy to setup for a sanity test, likely the OTA approach will be even easier.

Anyway, Foucault testing a well equilibriated smooth mirror for a 16" f/4.6 is easily done with more than sufficient accuracy, so long as you do it conscientiously:

Make a Couder mask with the proper number of openings (7 or 8). Read this and use the calculator.
Average your zonal readings if you're not sure.
Do this a couple of times over different days when you're fresh.
Maybe get somebody else to duplicate the readings if you've got doubts.

Best,
Mark

Edited by mark cowan (09/04/12 01:01 AM)

Post Extras:

Print Post

Remind Me!

Notify Moderator

ccaissie
sage

Reged: 09/13/10

Loc: Whitefield, Maine

Re: Refiguring 16" f/4.5 Meade new

new

[Re: mark cowan]
#5402675 - 09/04/12 10:22 AM

I actually prefer the pinstick method, finding it easier to get more consistent KE readings. Pins set in a rod using my Bridgeport gets them within a few thou.

Averaging readings...oh yeah! Like 4 sets.

I ignore the center zone out to 4" dia. I set the longitudinal scale by hitting the 32% zone pin, and setting the dial readout at .070" (Ignore the screenshot values in the early post...they don't match anything), and then take readings from there. I also take readings going in, by setting Zone 6 at a particular value and getting the relative distances. I can get it to +/- .005" pretty reliably, and with more practice, even better.

It will be easier to read when I strip off the coating.

Yes, Ronchi picks up zoniness and edge defects. This edge is pretty sharp, and I plan to do the edge-annulus-grind technique that I love so well, i.e. grind the face with 5 micron to get a .050 wide flat to get rid of the probable tiny turned edge (Texereau). I've heard others SCREAM at this technique, but it works for me.

I will setup the autocollimation with a 12.5 flat, already worked out in CAD what the lines will look like. I have used Peter J Smiths shearing interferometer, but how to hook that up in the OTA eyepiece is iffy.

I'm fond of the Bench Star test with autocollimation too, using an unlensed laser pointer with a really small diode...effectively as a point source.

Thanks all, this should be fun! (or why do it?)

Post Extras:

Print Post

Remind Me!

Notify Moderator

MKV
Carpal Tunnel

Reged: 01/20/11

Re: Refiguring 16" f/4.5 Meade new

new

[Re: ccaissie]
#5402799 - 09/04/12 11:31 AM

ccaissie, what you describe is precision, not accuracy. You are a meticulous worker and that shows in your repeated numbers. Averaging them out does nothing because the internal error is still present. Internal error simply means that just because your dial says it's 0.005" it doesn't mean it is 0.005". All mechanical devices have a certain inherent reading limit, a certain ± zone of uncertainty.

Let's assume your internal error is ± 0.005", and your readings are 1.05±0.005, 1.075±0.005 and 1.03±0.005. The average will be 1.052±0.005 - which means the true value will be some value between 1.047 and 1.057, not necessarily 1.052. To be on the safe side of the uncertainty, the value is 1.05.

Now, in some cases this is not significant, in others it is. Usually, in large, fast mirrors it is significant.

Take for example a spherometer whose radius is known to ±0.01 inch, and its sagittae readings are known to 0.001 inch. What is the internal error of the spherometer?

Well, assume you need one for your mirror, let's say of 5-inch radius. Your r = 5±0.01, and assume your sagitta s = 0.174±0.001 inches.

The result (ROC you obtain) will be no mor accurate than 98%. The ± uncertainty on the radius of curvature will be ~ 0.7 inches. So the ROC will be ~ 72±0.7" , or between 71.3 and 72.7 inches.

Does that effect the Foucualt readings on a 16-inch f/4.5 and the figure? I don't know, plug them in and see if they do or they don't.

There is plenty of literature on this subject of precision vs accuracy. A derivation of how the spherometer error was calculated is to be found in Advanced Telescope Making Techniques, Vol.2, by Willam Browne., p. 106.

Regards,

Mladen

Post Extras:

Print Post

Remind Me!

Notify Moderator

mark cowan
Vendor (Veritas Optics)

Reged: 06/03/05

Loc: salem, OR

Re: Refiguring 16" f/4.5 Meade new

new

[Re: ccaissie]
#5403145 - 09/04/12 02:41 PM

Quote:

Yes, Ronchi picks up zoniness and edge defects. This edge is pretty sharp, and I plan to do the edge-annulus-grind technique that I love so well, i.e. grind the face with 5 micron to get a .050 wide flat to get rid of the probable tiny turned edge (Texereau). I've heard others SCREAM at this technique, but it works for me.

This may be based on not actually knowing what you've got. There are no "probable" turned edges. Several methods can show you the edge condition and you should use one of those to determine it before adjusting for it. If you need more help just ask.

I've not used the pinstick method, but a word of warning - consistency is not the same as accuracy. Repeating bias will be consistent but inaccurate. Since you already described disagreement between two independent tests (star testing and Foucault) you need to get to accuracy first.

Oh yeah, though a laser diodes source is small, it's not round. But I've used one as the source for a 2D Hartman test with no problems. There's another independent test you could use, although making the mask is a *BLEEP*.

Have fun!

Best,
Mark

Post Extras:

Print Post

Remind Me!

Notify Moderator

Pinbout
Post Laureate

Reged: 02/22/10

Loc: Montclair

Re: Refiguring 16" f/4.5 Meade new

new

[Re: mark cowan]
#5403357 - 09/04/12 04:08 PM Attachment (33 downloads)

Quote:

Make a Couder mask with the proper number of openings (7 or 8). Read this and use the calculator.

you dont have to use a calculator.

Divide the radius by how ever many zones you want and the start drawing lines. works every time. [its a geometric construction of divide a circle into equal areas, machinery's handbook]

Post Extras:

Print Post

Remind Me!

Notify Moderator

mark cowan
Vendor (Veritas Optics)

Reged: 06/03/05

Loc: salem, OR

Re: Refiguring 16" f/4.5 Meade new

new

[Re: Pinbout]
#5403481 - 09/04/12 05:26 PM

That works too.

OP, measure the mask when you're done cutting for the actual zonal radii though if you use a geometric construction.

OTOH if you use calculation to construct a mask you can just use the calculated radii directly for the zones - the results in use are quite robust against small errors from mask construction.

Best,
Mark

Post Extras:

Print Post

Remind Me!

Notify Moderator

ccaissie
sage

Reged: 09/13/10

Loc: Whitefield, Maine

Re: Refiguring 16" f/4.5 Meade new

new

[Re: MKV]
#5407410 - 09/06/12 10:32 PM

Quote:

There is plenty of literature on this subject of precision vs accuracy. A derivation of how the spherometer error was calculated is to be found in Advanced Telescope Making Techniques, Vol.2, by Willam Browne., p. 106.

Absolutely one of may favorite articles, one that unlocked the use of calculus in my mind.

Well taken discussion. I will revive my caustic tester and use that later, as it uses dial indicators for readouts to .0001", and when I run my calcs I have assumed it is a conservative +/- .001". As I get smoother and closer, I may do the lateral readings to measure the caustic. I'd bet they are unreliable right now due to roughness.

Stripped the coating with nitric acid, easier to read the subtleties. I'm seeing via Ronchi and Foucault that the edge zones are irregular, as the shadows do not transit smoothly. My first exercise will be to smooth these with a 4-5" star lap, soft pitch and rouge, and watch for a smoothing of the ronchi lines, and a general improvement trend in reading of the figure. My previous figuring session years ago left a high zone at 90%, so my initial thinking will be to reduce and blend that. Will report in a few days.

Post Extras:

Print Post

Remind Me!

Notify Moderator

ccaissie
sage

Reged: 09/13/10

Loc: Whitefield, Maine

Re: Refiguring 16" f/4.5 Meade new

new

[Re: alnitak]
#5407424 - 09/06/12 10:37 PM

Yours was supposed to be the last..but hey, there was this incomplete 10" and I had to....

Post Extras:

Print Post

Remind Me!

Notify Moderator

Joe Cepleur
Pooh-Bah

Reged: 03/18/10

Loc: Dark North Woods

Re: Refiguring 16" f/4.5 Meade new

new

[Re: ccaissie]
#5407500 - 09/06/12 11:30 PM

Hate to confess, but I don't understand a word of this thread. I follow only that, lacking the ideal equipment for testing so large a mirror, ccaissie is determined to substitute tests that will still work perfectly well, and make the mirror right. Rock on!

Post Extras:

Print Post

Remind Me!

Notify Moderator

MKV
Carpal Tunnel

Reged: 01/20/11

Re: Refiguring 16" f/4.5 Meade new

new

[Re: ccaissie]
#5407595 - 09/07/12 12:35 AM

A caustic tester? Sweet. Now, with that you can definitely get a large fast mirror refigured to an accurate paraboloid. But it is time consuming and very sensitive to environment. Here is an excellent description of the process by someone who uses it professionally on a regular basis.

Caustic Test

Will be looking for your updates. Good luck!

Mladen

Post Extras:

Print Post

Remind Me!

Notify Moderator

Pages: 1 | 2 | 3 | 4 | (show all)

Extra information

11 registered and 16 anonymous users are browsing this forum.

Moderator: ausastronomer, richard7

Print Thread

Forum Permissions
      You cannot start new topics
      You cannot reply to topics
      HTML is disabled
      UBBCode is enabled

Thread views: 3821

Jump to

Cloudy Nights LLC
Cloudy Nights Sponsor: Astronomics