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

Need help with corrector plate and secondary tuning in C8

  • Please log in to reply
32 replies to this topic

#1 bluecoast

bluecoast

    Lift Off

  • *****
  • topic starter
  • Posts: 10
  • Joined: 13 Jul 2017

Posted 25 August 2019 - 09:40 PM

Hi all.

 

I have a couple of questions about the corrector plate and secondary in a C8.

 

A few years ago, when loosening the threaded retaining ring for Fastar installation, the entire secondary instead came loose inside the corrector plate hole. I had to remove the corrector plate to properly tighten the secondary. I also found NO shims -- none between secondary and corrector and none between corrector and OTA. I also found no marks I found meaningful on the corrector plate. I was very stupid and didn't add any clock marks between corrector plate and OTA, and didn't even mark the 'out' direction of the corrector. Even more stupid was that probably this could have gone back to Celestron as the root cause was the internal retaining ring on secondary assembly was loose and the scope was under warranty when this occurred. Again I acknowledge the stupidity.

 

Anyway I reassembled by centering the corrector in OTA as best I could and likewise the secondary in corrector as best I could, leaving uniform air gaps around each. I measured between the OTA and secondary with calipers as well. I wasn't 100% certain the corrector plate 'out' side was the same as original. I expected to reassemble and then have to do some trial and error fine tuning:

 

- clock position of secondary within corrector

- if things were uniformly bad, try flipping corrector so inner face is out

- if needed, try clocking corrector plate within OTA

 

When I put it back together and collimated though, it was actually pretty good. No worse than it was originally, in my estimation. So I left it as-is and have continued to use it for a couple more years. I couldn't help but think at the time though that:

 

- the optics are pretty uniformly made and spherical, so collimation will correct any alignment problem

- OR the OTA isn't super well tuned off production line, so reassembly without fine-tuning yields same performance as stock OR

- OR I got really lucky (doubtful)

 

My questions are:

1. Is it possible for me to fine-tune the corrector/secondary assembly/alignment to get better performance? What kind of tools/procedure/information would I need to reassemble with best possible alignment of primary, secondary, and corrector plate?

2. Can I do a test to verify which side of the corrector plate should be out? I don't have an optical flat, but I do have lasers, straightedges, and a granite surface plate.

3. Is there anything I can do to tune the primary? It would be nice to reduce focus backlash and tune grease/bushings etc. for a perfect sliding fit

 

also:

 

4. (maybe not related to above reassembly) I notice sometimes one one side of focus things tend to blur out vertically, and on the other side they blur horizonally. Collimating doesn't fix this. Other times the issue seems to be gone. I also have some chromatic fringes sometimes. Other times things are very sharp and crisp -- could this be differences in eyepieces? The primary mirror slopping around a bit? 

 

Just looking to discuss this with people smarter/more experienced than me -- feel free to chime in.

 

Thank you kindly!

 



#2 MKV

MKV

    Cosmos

  • *****
  • Posts: 8086
  • Joined: 20 Jan 2011
  • Loc: Florida, USA

Posted 25 August 2019 - 10:15 PM

The corrector can be tested using a water flat, by immersion. See this video by Ed Jones on Youtube for water flat testing.

 

https://www.youtube....h?v=TIAqTV2noJk 

 

Some correctors have 50% of correction on each side (i.e. there's no flat side). Otherwise, the (unofficial but logical) rule of thumb is that the flat side is facing the sky and the figured side is facing the primary.

 

Mladen



#3 davidc135

davidc135

    Viking 1

  • -----
  • Posts: 696
  • Joined: 28 May 2014
  • Loc: Wales, UK

Posted 26 August 2019 - 01:19 AM

I don't think it matters which side of the corrector is facing out but you might find a bit of glass that was flat enough to test it. Cut a piece of window glass into squares 21/2 ins on a side. Use a fluorescent bulb or, better, a monochromatic light source to check for Newton's fringes. A magnifying glass above the squares can help here. Find three squares of glass that only show a few fringes when combined with each other. The Schmidt curve will be very strongly concave in the outer part of the corrector- dozens of fringes if on the one side which they will be in Celestrons.

 

But it's not very important.   David


Edited by davidc135, 26 August 2019 - 01:25 AM.


#4 MKV

MKV

    Cosmos

  • *****
  • Posts: 8086
  • Joined: 20 Jan 2011
  • Loc: Florida, USA

Posted 26 August 2019 - 01:36 AM

I don't think it matters which side of the corrector is facing out 

Yes it does. If the flat side is facing the primary you're likely to get ghost reflections at the focal plane unless you intentionally tilt the corrector a couple of degrees.     



#5 davidc135

davidc135

    Viking 1

  • -----
  • Posts: 696
  • Joined: 28 May 2014
  • Loc: Wales, UK

Posted 26 August 2019 - 03:37 AM

Yes it does. If the flat side is facing the primary you're likely to get ghost reflections at the focal plane unless you intentionally tilt the corrector a couple of degrees.     

Ah, OK. But have they been shown in practise?

 

Scene A; reflection first off the 'flat' side facing the primary then second reflection off the Schmidt surface to eventually form the ghost.

Scene B; reflection first off the Schmidt curve facing the primary followed by the second reflection off the flat side to eventually form the ghost.

Difference?

 

Or is there another mechanism? I'm not making a claim here so much as asking the question.

 

David



#6 MKV

MKV

    Cosmos

  • *****
  • Posts: 8086
  • Joined: 20 Jan 2011
  • Loc: Florida, USA

Posted 26 August 2019 - 07:55 AM

David,

 

It can be shown by raytracing that a flat side will result in ghost images. This is a well established fact. All you have to do is search the subject and you'll find tons of articles to support this. I could post a raytrace analysis as well but that's an overkill. I have other things to do.

 

The convex (figured) side of the corrector faces the mirror to avoid ghosting.

 

If for some reason the corrector is mounted with the flat side facing the mirror, it can be tilted by 1 to 2 degrees with respetc to the optical axis withuot any harm.  This is precisely what was done in this NASA paper

 

http://adsabs.harvar...BAA..103..294M 

 

In this exaggerated drawing of the system in question, you can see a no-to-scale (for better clarity) drawing of the the titlted corrected (1.5 degrees).

 

http://articles.adsa...0&filetype=.gif

 

I found at least a dozen other articles on this subject in  about five minutes in Chrome.  Please research the topic first.  

 

Regards,

Mladen



#7 davidc135

davidc135

    Viking 1

  • -----
  • Posts: 696
  • Joined: 28 May 2014
  • Loc: Wales, UK

Posted 26 August 2019 - 08:59 AM

Mladen. It's interesting. I'll look into it. Anybody fancy an experiment?

 

Bluecoast. Hopefully others will address the other practical points you raised. Probably not a good idea for me to devote too much of this thread to ghost questions, interesting though they are.

 

David


Edited by davidc135, 26 August 2019 - 09:06 AM.


#8 GShaffer

GShaffer

    Knight of Ni

  • *****
  • Moderators
  • Posts: 6397
  • Joined: 28 Feb 2009
  • Loc: Bogart, Ga USA

Posted 26 August 2019 - 09:36 AM

Have good performing C11.....several years ago I bought another for $200 that the owner said had issues. After investigation all that was wrong with it was the corrector plate had been put on backwards. once we turned it around and did some testing and turning of it I ended up with another C11 that performed as well as the one I already had. The repaired one was a carbon fiber shell and took longer to be ready to use thermally so I sold it off.....So I can say without a shadow of doubt a reversed corrector definitely impacts performance.


  • davidc135 and wrnchhead like this

#9 davidc135

davidc135

    Viking 1

  • -----
  • Posts: 696
  • Joined: 28 May 2014
  • Loc: Wales, UK

Posted 26 August 2019 - 10:07 AM

My apologies Bluecoast but it's hard to bite my lip. In a similar thread July '17 by stargeeser, Rod Mollise and DavidG both say it doesn't make any of difference which side is out.   David


Edited by davidc135, 26 August 2019 - 10:08 AM.


#10 MKV

MKV

    Cosmos

  • *****
  • Posts: 8086
  • Joined: 20 Jan 2011
  • Loc: Florida, USA

Posted 26 August 2019 - 10:40 AM

My apologies Bluecoast but it's hard to bite my lip. In a similar thread July '17 by stargeeser, Rod Mollise and DavidG both say it doesn't make any of difference which side is out.   David

I find it interesting that someone will doubt a NASA paper but accept somewhat anecdotal and possibly incomplete evidence to the contrary by amateurs. confused1.gif  Incidentally, when quoting someone it helps to offer a traceable reference to the source.

 

As I said, a ghost image analysis in a professional raytrace program is all that's needed to corroborate what's in the NASA article, even if I initially doubted it.. The software normally used by most ATMS is OSLO.edu, which is an excellent freebie version of a professional program under the same name but different extension.

 

Unfortunately, the .edu version doesn't support ghosting analysis (that choice is disabled), so I imagine most ATMs who use it cannot actor ghosting as an impediment in their assessment.  And this is not an issue only with Schmidt corrector plates but with air spaced lens lenses as well, especially with large gaps and concave surfaces.

 

Speaking of Dave, I believe DavidG would be the first to tell you that the flat side of the corrector shouldn't be flat! :o). I know David, and he usually doesn't miss anything.

 

Regards,

Mladen



#11 bluecoast

bluecoast

    Lift Off

  • *****
  • topic starter
  • Posts: 10
  • Joined: 13 Jul 2017

Posted 26 August 2019 - 10:50 AM

Is there a consensus among experts on how the corrector is oriented from the factory? Flat face out?

 

If it's possible for me to determine the flat side, and it's also known which way they're oriented from the factory, I'll try to return it to that orientation regardless of what it is (this is my diplomatic way of not entering the ghost image debate).

 

Is there any better method than trial-and-error for clocking the secondary relative to the primary? 

 

My common-sense feeling (which could be wrong and is based on no real experience at all) would be:

 

- corrector clock position relative to primary doesn't matter (corrector curvature very slight and guaranteed uniform by manufacture method, e.g. grinding flat side while corrector held distorted in fixture)

- corrector concentricity to primary may be very important

- secondary clock position relative to primary MAY be important if they are a matched pair (e.g. secondary figured to compensate for flaw in primary) -- not clear whether hand-tuning or matching is done on mass-produced optics anymore

- secondary concentricity to primary not very important (assuming above factors good) as this would be corrected by collimation

 

Am I wrong about any or all of these?



#12 davidc135

davidc135

    Viking 1

  • -----
  • Posts: 696
  • Joined: 28 May 2014
  • Loc: Wales, UK

Posted 26 August 2019 - 11:15 AM

Regardless of the ghost debate it would be as well to put the flat side out... you're not going to lose anything. Still, if the flat faces in, it could be an opportunity to have a look at a bright star re ghosts... in the spirit of enquiry.

 

The supposedly flat side may not be exactly flat or even an exact figure of revolution. David



#13 Vla

Vla

    Messenger

  • *****
  • Posts: 491
  • Joined: 08 Mar 2014

Posted 26 August 2019 - 12:01 PM

I don't think either side is flat. Johnson's patent on making Schmidt corrector with the vacuum method shows doing both sides, and there are good reason for it. The deeper Schmidt curve, the more hard is to get it accurately. Anyway, raytracing shows no difference whatsoever from reversing one-sided corrector, and I failed to find any specific reference that it matters for ghosting either.



#14 Vla

Vla

    Messenger

  • *****
  • Posts: 491
  • Joined: 08 Mar 2014

Posted 26 August 2019 - 12:04 PM

David,

 

It can be shown by raytracing that a flat side will result in ghost images. This is a well established fact. All you have to do is search the subject and you'll find tons of articles to support this. I could post a raytrace analysis as well but that's an overkill. I have other things to do.

 

The convex (figured) side of the corrector faces the mirror to avoid ghosting.

 

If for some reason the corrector is mounted with the flat side facing the mirror, it can be tilted by 1 to 2 degrees with respetc to the optical axis withuot any harm.  This is precisely what was done in this NASA paper

 

http://adsabs.harvar...BAA..103..294M 

 

In this exaggerated drawing of the system in question, you can see a no-to-scale (for better clarity) drawing of the the titlted corrected (1.5 degrees).

 

http://articles.adsa...0&filetype=.gif

 

I found at least a dozen other articles on this subject in  about five minutes in Chrome.  Please research the topic first.  

 

Regards,

Mladen

Mladen, that paper does not mention tilting Schmidt corrector to avoid ghosting. The illustration merely shows the corrector had a hinged carrier, so that it can be flipped out of the way to switch from one configuration to the other. Tilting Schmidt corrector inevitably induces all-field coma (some small in comparison astigmatism and spherical aberration as well).


Edited by Vla, 26 August 2019 - 01:30 PM.


#15 davidc135

davidc135

    Viking 1

  • -----
  • Posts: 696
  • Joined: 28 May 2014
  • Loc: Wales, UK

Posted 26 August 2019 - 03:47 PM

Bluecoast, You mentioned that you have a granite surface plate. Are they supplied polished or fine ground? If you can get interference fringes the Schmidt curve will jump out. It's so deep and the shape so characteristic. Or even the diagonal surface of a porro prism.   David



#16 davidc135

davidc135

    Viking 1

  • -----
  • Posts: 696
  • Joined: 28 May 2014
  • Loc: Wales, UK

Posted 26 August 2019 - 04:04 PM

Is there a consensus among experts on how the corrector is oriented from the factory? Flat face out?

 

If it's possible for me to determine the flat side, and it's also known which way they're oriented from the factory, I'll try to return it to that orientation regardless of what it is (this is my diplomatic way of not entering the ghost image debate).

 

Is there any better method than trial-and-error for clocking the secondary relative to the primary? 

 

My common-sense feeling (which could be wrong and is based on no real experience at all) would be:

 

- corrector clock position relative to primary doesn't matter (corrector curvature very slight and guaranteed uniform by manufacture method, e.g. grinding flat side while corrector held distorted in fixture)

- corrector concentricity to primary may be very important

- secondary clock position relative to primary MAY be important if they are a matched pair (e.g. secondary figured to compensate for flaw in primary) -- not clear whether hand-tuning or matching is done on mass-produced optics anymore

- secondary concentricity to primary not very important (assuming above factors good) as this would be corrected by collimation

 

Am I wrong about any or all of these?

 

Your common sense feelings are sound Imo, the secondary may be slightly aspheric for the reason you give. But also I don't think the corrector is dependably uniform or free from astigmatism etc and so its rotation may be necessary. It may well be trial and error but, hopefully, others with practical experience can be more specific.

 

I'd be surprised if the Schmidt asphericity was on both sides in Celestron's case.

 

David



#17 DAVIDG

DAVIDG

    Cosmos

  • *****
  • Posts: 8985
  • Joined: 02 Dec 2004
  • Loc: Hockessin, De

Posted 26 August 2019 - 04:11 PM

I don't think either side is flat. Johnson's patent on making Schmidt corrector with the vacuum method shows doing both sides, and there are good reason for it. The deeper Schmidt curve, the more hard is to get it accurately. Anyway, raytracing shows no difference whatsoever from reversing one-sided corrector, and I failed to find any specific reference that it matters for ghosting either.

 Celestron put  the correction on both sides of the  corrector only for their Schmidt cameras which required more correction then their  telescopes. In the telescopes the correction is only on one side but the "flat" surface is many waves from flat since the  correctors are made from float glass and the "flat" side is how the glass is received.   

  As pointed out in other threads here were we raytraced SCT it makes no difference optically what side of the corrector faces the sky. There is only a very slight change in the position of the focal plane because of the thickness of the corrector depending on what side is sky facing. Also most modern SCT have the corrector AR coated on both sides so that would reduce any ghosting.

 

                    - Dave   


Edited by DAVIDG, 26 August 2019 - 04:17 PM.


#18 Vla

Vla

    Messenger

  • *****
  • Posts: 491
  • Joined: 08 Mar 2014

Posted 26 August 2019 - 07:20 PM

 Celestron put  the correction on both sides of the  corrector only for their Schmidt cameras which required more correction then their  telescopes. In the telescopes the correction is only on one side but the "flat" surface is many waves from flat since the  correctors are made from float glass and the "flat" side is how the glass is received.   

  As pointed out in other threads here were we raytraced SCT it makes no difference optically what side of the corrector faces the sky. There is only a very slight change in the position of the focal plane because of the thickness of the corrector depending on what side is sky facing. Also most modern SCT have the corrector AR coated on both sides so that would reduce any ghosting.

Not sure about one-sided profile. Still have some old Meade catalog in which they implicitly say that they SCT's use double profile. Besides, there is little difference between f/2 Schmidt camera and the standard f/2/10 SCT, which needs equivalent of about f/2.2 Schmidt camera corrector. The camera is generally less demanding then the telescope, so what would be the reason?

 

As for the change in best focus location due to flipping the corrector around, it is a couple of 1/000 of mm. Not worth mentioning.



#19 davidc135

davidc135

    Viking 1

  • -----
  • Posts: 696
  • Joined: 28 May 2014
  • Loc: Wales, UK

Posted 27 August 2019 - 02:00 AM

Meade divided the correction between the two surfaces, I believe, using Schmidt's vacuum pan method. Perhaps for legal reasons.

 

Didn't the Celestron cameras work at F/1.5 or F/1.65 according to aperture?  If at F/1.65 that would be (2.2/1.65) cubed = x2.36 the correction and also 2.36 x stress, if the same glass thickness was used. So quite a jump and perhaps too great a risk of breakage even with the solid support their method used. And drastically thinning the glass corrector to reduce stresses may not have been practical.

 

Dave has an interferogram of a Celestron sct corrector 'flat' surface somewhere. I'm sure I've seen it? Gently wavy, IIRC.

 

David


Edited by davidc135, 27 August 2019 - 07:39 AM.


#20 Vla

Vla

    Messenger

  • *****
  • Posts: 491
  • Joined: 08 Mar 2014

Posted 27 August 2019 - 07:43 AM

If the Schmidt camera is f/1.5 that is certainly good reason to go double side (more than 3 times deeper than at f/2.2). What kind of "wavy" was that interferogram?



#21 davidc135

davidc135

    Viking 1

  • -----
  • Posts: 696
  • Joined: 28 May 2014
  • Loc: Wales, UK

Posted 27 August 2019 - 07:58 AM

I happened on it in some past thread IIRC but can't remember where. Hopefully Dave has a record. With it I could compare Celestron's superior selected discs of float glass with those used by Criterion where the unworked 'flat' side was pulled by vacuum against a profiled back plate. The Criterion surface could be tens of waves irregular as against a few waves in the Celestron case which were also irregular, random. Not wavy as in Schmidt curve wavy. But this from memory.

 

David


Edited by davidc135, 27 August 2019 - 08:00 AM.


#22 MKV

MKV

    Cosmos

  • *****
  • Posts: 8086
  • Joined: 20 Jan 2011
  • Loc: Florida, USA

Posted 27 August 2019 - 08:55 AM

As pointed out in other threads here were we raytraced SCT it makes no difference optically what side of the corrector faces the sky. There is only a very slight change in the position of the focal plane because of the thickness of the corrector depending on what side is sky facing. Also most modern SCT have the corrector AR coated on both sides so that would reduce any ghosting.

The only reason I have seen mentioned why the flat side should be facing the mirror is to avoid ghost images. Other than that, it makes no difference optically, since the lens is thin and has very little refractive power.

 

FYI, ghosting is frequently reported with Schmidt systems, both cameras and SCTs. and always involves bright objects. This can happen even on curved surfaces due to minute chips in the lens, which then become tiny point sources of light and are imaged parasitically by the system.

 

Mladen



#23 evan9162

evan9162

    Surveyor 1

  • *****
  • Posts: 1605
  • Joined: 20 Sep 2013
  • Loc: ID

Posted 27 August 2019 - 09:48 AM

Not sure about one-sided profile. Still have some old Meade catalog in which they implicitly say that they SCT's use double profile. Besides, there is little difference between f/2 Schmidt camera and the standard f/2/10 SCT, which needs equivalent of about f/2.2 Schmidt camera corrector. The camera is generally less demanding then the telescope, so what would be the reason?

 

As for the change in best focus location due to flipping the corrector around, it is a couple of 1/000 of mm. Not worth mentioning.

I can confirm that, at least for old Meades, the curve is ground into both sides of the corrector.  My 1980's era 10" corrector plate shows the curve on both sides when doing contact fringe testing with an optical flat.

 

However, the profiles are not coaxial to each other - one side is significantly offset from the other  (nearly 10mm) - certainly the result of hurried manufacturing during the Halley era. 



#24 MKV

MKV

    Cosmos

  • *****
  • Posts: 8086
  • Joined: 20 Jan 2011
  • Loc: Florida, USA

Posted 27 August 2019 - 10:11 AM

However, the profiles are not coaxial to each other - one side is significantly offset from the other  (nearly 10mm) - certainly the result of hurried manufacturing during the Halley era. 

Woooow! 



#25 wrnchhead

wrnchhead

    Viking 1

  • *****
  • Posts: 937
  • Joined: 28 Aug 2017
  • Loc: NE Kansas

Posted 27 August 2019 - 11:04 AM

This is all very interesting. I would have never even guessed that something like this could be a point of contention, I always thought they all used the same basic optical design. Surprised to hear various experienced experts with completely different conclusions. 




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