Jump to content


Photo

Caustic vs. Pinstick

  • Please log in to reply
53 replies to this topic

#1 ccaissie

ccaissie

    Viking 1

  • -----
  • Posts: 634
  • Joined: 13 Sep 2010
  • Loc: Whitefield, Maine

Posted 19 March 2013 - 07:43 AM

Bummed. I got a mirror test and it doesn't jive with all the work I did using "Foucault Test Analysis V2.08".

Can someone just review my numbers and see where I went 1/3 wave off?

Pin location /Moving slitless readings
2.50/.071
4.36/.115
5.62/.159
6.63/.205
7.13/.226
7.59/.250

I had used the star test to correlate my figuring earlier, and when I saw it had some discrepancies, I doubted my results, and started relying on the easier method of figuring to the above program.

I started using Figure XP, and I get the same results...should have been an exceptional mirror. See picture.

I set up my Caustic tester, and it's showing the error. Closer star testing does show error.

So why are these Foucault programs giving me bad numbers?
Could I have misinterpreted the term "effective radius"? Isn't that the center of a mask aperture or exact pin location?

I must be a dummy. Calculating r^2/2R gives me about 80% smaller readings. Why didn't I just do it the old way and skip the software?

Attached Files



#2 ccaissie

ccaissie

    Viking 1

  • -----
  • Posts: 634
  • Joined: 13 Sep 2010
  • Loc: Whitefield, Maine

Posted 19 March 2013 - 08:14 AM

Here's the input values.

Attached Files



#3 Pinbout

Pinbout

    Cosmos

  • *****
  • Posts: 8013
  • Joined: 22 Feb 2010
  • Loc: nj

Posted 19 March 2013 - 08:38 AM

and your using effective radius as opposed to inside/outside?

this is what I got from you numbers.

Attached Files



#4 Pinbout

Pinbout

    Cosmos

  • *****
  • Posts: 8013
  • Joined: 22 Feb 2010
  • Loc: nj

Posted 19 March 2013 - 08:45 AM

and

Attached Files



#5 Mark Harry

Mark Harry

    Vendor

  • *****
  • Vendors
  • Posts: 6188
  • Joined: 05 Sep 2005
  • Loc: Northeast USA

Posted 19 March 2013 - 09:49 AM

You can verify the transverse abberation on a calculator. If it doesn't compare favorably with XP's result, I can understand the possible discrepancy. A transverse error of 8.0 is around 1/.3-something wave, roughly 3 waves total error. fwiw,
M.

#6 ccaissie

ccaissie

    Viking 1

  • -----
  • Posts: 634
  • Joined: 13 Sep 2010
  • Loc: Whitefield, Maine

Posted 19 March 2013 - 09:50 AM

Right. I used "effective radius" setting in the XP program, and plugged in my numbers.

In the Foucault v2.08 I specified "FoucaultTest (Pin Stick)", and plugged in my numbers.

With a pin, there's no inside/outside measure. I like the process, and am convinced it's an accurate way to measure the zones. If I had hand calculated the KE values, and just figured from that I'd have been OK. Bungled somewhere in using the software.

Where? Should have persevered and trusted the star test, but shlepping the mirror in & out was tedious. It's a beautiful mirror, just wrong.

Now that I've got the ol' Caustic rig out, I'm using that from now on. For this checkup, I used the poorman's version which showed the overcorrection generally. When I get it set up, I'll use the official mask and statistical measurements.

#7 ccaissie

ccaissie

    Viking 1

  • -----
  • Posts: 634
  • Joined: 13 Sep 2010
  • Loc: Whitefield, Maine

Posted 19 March 2013 - 09:55 AM

and your using effective radius as opposed to inside/outside?
this is what I got from you numbers.

Look what your version did to the mask/pin locations? Not at all what I used.

C

#8 ccaissie

ccaissie

    Viking 1

  • -----
  • Posts: 634
  • Joined: 13 Sep 2010
  • Loc: Whitefield, Maine

Posted 19 March 2013 - 10:00 AM

I don't use the transverse values function, at least in cases of such great error.

What's wrong with using wavefront analysis, at this stage at least?

Why are there such discrepancies in the software values I got and the actual ideal numbers?

What box did I fail to click? That's where I'm at now.

#9 Pinbout

Pinbout

    Cosmos

  • *****
  • Posts: 8013
  • Joined: 22 Feb 2010
  • Loc: nj

Posted 19 March 2013 - 10:04 AM

I know I'm trying to make it look worse. :p

#10 ccaissie

ccaissie

    Viking 1

  • -----
  • Posts: 634
  • Joined: 13 Sep 2010
  • Loc: Whitefield, Maine

Posted 19 March 2013 - 10:11 AM

Worse? Oy gevalt!
I figured a really fine 12.0" f/3.0 without software ...pinstick and autocollimation. Live and learn. :bawling:

#11 Pinbout

Pinbout

    Cosmos

  • *****
  • Posts: 8013
  • Joined: 22 Feb 2010
  • Loc: nj

Posted 19 March 2013 - 10:36 AM

Post deleted by Pinbout

#12 Pinbout

Pinbout

    Cosmos

  • *****
  • Posts: 8013
  • Joined: 22 Feb 2010
  • Loc: nj

Posted 19 March 2013 - 10:51 AM

when I enter your numbers it seems way off when I hit fit best conic, it makes the graph for -.5 conic, it looks like your really good.

#13 ccaissie

ccaissie

    Viking 1

  • -----
  • Posts: 634
  • Joined: 13 Sep 2010
  • Loc: Whitefield, Maine

Posted 19 March 2013 - 11:32 AM

Yeah, I get a -.5 too, if I set the KE as fixed.

No, 3 software programs (XP, v2.08, and Zone.exe) I have tried give me great results with the numbers I plugged in.

It must be something about the radii measurements, no?
But when I use a mask setup in XP, and set the inner and outer at +/- 0.1" my pinstick values, it gives an effective radius of my pinstick value, confirming that the effective radius is the mean of the mask aperture radii... and the same fine excellent wave readout.

If I compute zones by r^2/2R, I get different numbers, smaller.


All 3 programs are happy with a 7.62 radius... my zone 6, giving a KE of .250" with moving source.

So, I use the pinstick setting in the software, or effective radii, and specify the exact location of the pins.

I find that I get the same averaged measurements whether I call the setting at the first pair of shadows hitting the pin, or the mid-gray zone, or visualizing the crest in 3-D, or by calling it as the last bit of light crosses the pin.

The attached photo is not great, the crest on the right is not set right, and there's vignetting, but you get the idea of what I mean. just a .001 change in the KE setting makes a difference in how the shadow crest hits the pin. so that's OK, I think...averaged measurements and all that.

Something I did very wrong, related to software assumptions, and it's making me want to can the software route and go back to my antediluvian methods. My mirrors made back then are still excellent.

Attached Files



#14 Mike Lockwood

Mike Lockwood

    Vendor, Lockwood Custom Optics

  • *****
  • Vendors
  • Posts: 478
  • Joined: 01 Oct 2007
  • Loc: Usually in my optical shop

Posted 19 March 2013 - 11:42 AM

Isn't that the center of a mask aperture or exact pin location?

Yes, you are correct for the case of a pin stick.

I must be a dummy. Calculating r^2/2R gives me about 80% smaller readings. Why didn't I just do it the old way and skip the software?

Using your numbers and r^2/2R (moving source) I get the same thing as FigureXP for the outer two zones:

(7.59^2/285.4) - (2.5^2/285.4) = 0.20185 - 0.0219 = 0.180"
(7.13^2/285.4) - (2.5^2/285.4) = 0.17812 - 0.0219 = 0.156"

Not sure what error your caustic or star testing is showing.

#15 Mark Harry

Mark Harry

    Vendor

  • *****
  • Vendors
  • Posts: 6188
  • Joined: 05 Sep 2005
  • Loc: Northeast USA

Posted 19 March 2013 - 11:53 AM

If Danny's ML graph is correct, it's way undercorrected in the outer half of the mirror, with a hole (in the center)
M.

#16 Pinbout

Pinbout

    Cosmos

  • *****
  • Posts: 8013
  • Joined: 22 Feb 2010
  • Loc: nj

Posted 19 March 2013 - 12:02 PM

no my ml graph is for the wrong zonal measurements.
but when I use his radii its even worse.

#17 ccaissie

ccaissie

    Viking 1

  • -----
  • Posts: 634
  • Joined: 13 Sep 2010
  • Loc: Whitefield, Maine

Posted 19 March 2013 - 12:26 PM

"Using your numbers and r^2/2R (moving source) I get the same thing as FigureXP for the outer two zones:

(7.59^2/285.4) - (2.5^2/285.4) = 0.20185 - 0.0219 = 0.180"
(7.13^2/285.4) - (2.5^2/285.4) = 0.17812 - 0.0219 = 0.156"

So, why do my entered measurements of .250" and .226", respectively, in my 3 programs result in such a marvelous wavefront chart?

My Caustic measurements were rough, (x= +/- .0005), but confirmed significant overcorrection. Star test showed earlier secondary shadow breakout outside focus at 540X.

I'm missing something here in the software.

#18 Mike Lockwood

Mike Lockwood

    Vendor, Lockwood Custom Optics

  • *****
  • Vendors
  • Posts: 478
  • Joined: 01 Oct 2007
  • Loc: Usually in my optical shop

Posted 19 March 2013 - 12:44 PM

So, why do my entered measurements of .250" and .226", respectively, in my 3 programs result in such a marvelous wavefront chart?

Those are the ideal readings according to three programs AND by the classical/proper formula - I don't understand why the wavefront shouldn't look good!?

My Caustic measurements were rough, (x= +/- .0005), but confirmed significant overcorrection. Star test showed earlier secondary shadow breakout outside focus at 540X.
I'm missing something here in the software.

What is significant?

Accurately-figured mirrors usually show overcorrection in the star test because the mirror is usually cooling. This is why lab tests are often more accurate than star testing.

I would recommend testing the mirror with two different masks and comparing the results with the pinstick. If they all agree to reasonable accuracy, I'd be suspicious of the star testing results.

#19 Pinbout

Pinbout

    Cosmos

  • *****
  • Posts: 8013
  • Joined: 22 Feb 2010
  • Loc: nj

Posted 19 March 2013 - 12:54 PM

ok I finally inputted your numbers correctly and got the same result in figurexp.

as i get older I become more adhd. :lol:

#20 ccaissie

ccaissie

    Viking 1

  • -----
  • Posts: 634
  • Joined: 13 Sep 2010
  • Loc: Whitefield, Maine

Posted 19 March 2013 - 02:47 PM

"If they all agree to reasonable accuracy, I'd be suspicious of the star testing results."

Or the coater's 1/3 wave .73 strehl test? Will this ever end?

Made a Couder screen, tested it, 4 sets of measurements...overcorrected by 1/5.4 p-v.

I star test at various times...8 p.m., 5 a.m. so temps are not always cooling, but you're right, I've seen most of my scopes go a bit over while cooling. Except my 6" f/11 sphere which gets perfect.

I'm fixing my Caustic setup and will get a good set of numbers from that.

Thanks all for the feedback. I was hoping that someone would say, "oh yeah, these 3 programs can't calculate pinstick setups...use paper and pencil only."

Obviously, I don't know how to put the numbers into XP correctly because of the discrepancy in the "ideal figures" column, yet got great wave shape. Screwed up on v2.0B, and ZONE.exe is wrong somehow.

Flummoxed.

#21 Mike Lockwood

Mike Lockwood

    Vendor, Lockwood Custom Optics

  • *****
  • Vendors
  • Posts: 478
  • Joined: 01 Oct 2007
  • Loc: Usually in my optical shop

Posted 19 March 2013 - 03:23 PM

Made a Couder screen, tested it, 4 sets of measurements...overcorrected by 1/5.4 p-v.

Evaluated with a program, or with a calculator? How does the figure look when using FigureXP?

Thanks all for the feedback. I was hoping that someone would say, "oh yeah, these 3 programs can't calculate pinstick setups...use paper and pencil only."

My experience shows that they work with pinsticks.....

Obviously, I don't know how to put the numbers into XP correctly because of the discrepancy in the "ideal figures" column, yet got great wave shape. Screwed up on v2.0B, and ZONE.exe is wrong somehow.
Flummoxed.

I'd like to see the comparison of the Couder mask data to the pinstick data using FigureXP, and how all of that compares to your calculations.

#22 Dave O

Dave O

    Ranger 4

  • *****
  • Posts: 380
  • Joined: 21 Dec 2011
  • Loc: Sri Lanka

Posted 19 March 2013 - 10:33 PM

"If they all agree to reasonable accuracy, I'd be suspicious of the star testing results."

Or the coater's 1/3 wave .73 strehl test? Will this ever end?


Perhaps you should have it tested by a different optician? If your testing indicates it is a good figure, I would not rush into changing anything until I was sure it needed changing (and what the necessary changes are).

#23 ccaissie

ccaissie

    Viking 1

  • -----
  • Posts: 634
  • Joined: 13 Sep 2010
  • Loc: Whitefield, Maine

Posted 20 March 2013 - 09:34 AM

Yes. I'll sit on it for a while, more thinking than anything, and when I'm finally convinced, I'll act.

Supposed to be a hobby, not an obsession.

#24 Pinbout

Pinbout

    Cosmos

  • *****
  • Posts: 8013
  • Joined: 22 Feb 2010
  • Loc: nj

Posted 20 March 2013 - 10:51 AM

Supposed to be a hobby, not an obsession.


yeah but an issue like this is one that really gnaws at you, or at least me. :foreheadslap:

#25 Norm Meyer

Norm Meyer

    Mariner 2

  • *****
  • Posts: 252
  • Joined: 08 Feb 2009
  • Loc: Warren, ME 04864

Posted 20 March 2013 - 11:26 AM

That's why I like spherical systems when possible or
null testing. I guess I'm lazy.

Norm






Cloudy Nights LLC
Cloudy Nights Sponsor: Astronomics