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

Post Script: Improving the wavefront quality of the C14 Edge

  • Please log in to reply
79 replies to this topic

#26 jhayes_tucson

jhayes_tucson

    Fly Me to the Moon

  • *****
  • topic starter
  • Posts: 7099
  • Joined: 26 Aug 2012
  • Loc: Bend, OR

Posted 03 May 2017 - 11:45 PM

Thanks John, I just meant to show someone else confirmed your findings on a different model. So it seems this is a common thing.

 

 

 

 

Tolga,

That is completely right.  I don't claim to be the first to do this by any means.  There are some folks in Europe who have shown similar interferometric results on Celestron scopes and that's why I decided to try it.

John



#27 tolgagumus

tolgagumus

    Vendor

  • *****
  • Vendors
  • Posts: 3630
  • Joined: 21 Jul 2014
  • Loc: NJ

Posted 04 May 2017 - 12:18 AM


Thanks John, I just meant to show someone else confirmed your findings on a different model. So it seems this is a common thing.



Tolga,
That is completely right. I don't claim to be the first to do this by any means. There are some folks in Europe who have shown similar interferometric results on Celestron scopes and that's why I decided to try it.
John

John I'm sorry. By no means I was trying to say you are not the first one. My point was this looks like a common sct problem. As you can see in the video, it was a different brand.

#28 SandyHouTex

SandyHouTex

    Skylab

  • *****
  • Posts: 4004
  • Joined: 02 Jun 2009
  • Loc: Houston, Texas, USA

Posted 04 May 2017 - 08:53 AM

 

 

Wow, this is incredible!  I wonder if even the synta manufacturer knows about this.

Last I heard, the C14s are assembled in Torrance:

 

http://www.chuckhawk..._connection.htm

 

Yes, that is true; but all of the components except the corrector plate come from China (AKA Synta.)  The primary is fully manufactured and qualified in China.  The secondaries are delivered as nominally spherical; but as many have reported here, the secondaries are reworked in Torrance to minimize the total wavefront error.   There is a lot of work done in Torrance but it's an effort split between China and the US to deliver a finished C14.  Many of the other scopes (of all other sizes) still go through Torrance as well.

 

John

 

Yes, that is what the linked article says.

 

My point was, that being assembled in Torrance, that would be where the clocking of the corrector and secondary would occur for optimum performance.



#29 rmollise

rmollise

    ISS

  • *****
  • Posts: 22874
  • Joined: 06 Jul 2007

Posted 04 May 2017 - 09:44 AM

This is a postscript to my testing article found here:  https://www.cloudyni...elescopes-r3095

 

A while back, I posted an article about testing the two C14 Edge scopes that I have in my shop.   The first scope that I tested is the one that I most often use for imaging and it tested at a Strehl of 0.845 (0.065 waves rms.)  Anything over 0.8 meets the MarĂ©chal criteria for a diffraction limited system, so that's not a bad system.  Still a number of folks were surprised that the system didn't measure better.

 

Since I did the original test, I've installed a carbon fiber tube on the telescope so I scheduled another session on the interferometer, which I just finished today.  Along the way, I've seen a number of articles showing that it is possible to improve the wavefront quality of Celestron systems by simply rotating the corrector plate to minimize wavefront errors.  That seemed like a good thing to try since astigmatism was the predominant wavefront error.  I also knew that the secondary was nearly a perfect sphere (tested in another session on a WYKO 6000.)   I was able to get the telescope aligned fairly quickly and positioned the field point to be within about +/-0.5 mm from center of the field.  I used all of the same components as in my original report so I didn't go through all of the calibration steps that I did the first time around.  

 

I then rotated the corrector plate assembly by~ 90 degrees and realigned the secondary.  The wavefront improved dramatically, and I could see that it might do better with a bit more rotation.  So I rotated the corrector assembly by another ~15 degrees.  Measurements showed that the astigmatism term was reduced by more than half to about 1/20 wave.  I was able to dial the coma term down to 0.010 waves and the total wavefront error decreased to 0.235 waves PVq(99%)  [meaning that 99% of the data fell within a range of 0.235 waves] with an rms of only 0.043 waves and a Strehl of 0.93!  The "measurement" was actually an average of 164 phase measurements to average out air turbulence.  I've attached the "shop screen" display from the 4D Foresight software.  This level of wavefront performance on a 14" telescope is exceptional--regardless of the manufacturer.

 

This result clearly implies that the clock alignment marks on the Celestron components are not very accurate.  I think that this may occur either because the sensitivity of the test used to place the marks is not very sensitive, or that there is a systematic error in the procedure.

 

I didn't try to realign the second scope but I suspect that it might benefit from a similar "re-alignment."

 

John

Back in the Day, I recall Celestron's test setup allow them to rotate the corrector. It sounds as if Synta dispensed with that part of the procedure.



#30 jhayes_tucson

jhayes_tucson

    Fly Me to the Moon

  • *****
  • topic starter
  • Posts: 7099
  • Joined: 26 Aug 2012
  • Loc: Bend, OR

Posted 04 May 2017 - 10:50 AM

Back in the Day, I recall Celestron's test setup allow them to rotate the corrector. It sounds as if Synta dispensed with that part of the procedure.

 

 

 

I believe that they still try to optimize the system by centering the corrector plate.  As I have pointed out in other threads, the system is very insensitive to small decenter errors (~ < +/- 3 mm) so I believe that this step is a waste of time.  I don't know what they used to do in terms of clock angle adjustment on the corrector assembly but if they aren't currently doing it, this data suggests that they should be.

 

John


Edited by jhayes_tucson, 04 May 2017 - 10:51 AM.

  • Gene@CN likes this

#31 rmollise

rmollise

    ISS

  • *****
  • Posts: 22874
  • Joined: 06 Jul 2007

Posted 04 May 2017 - 04:18 PM

 

Back in the Day, I recall Celestron's test setup allow them to rotate the corrector. It sounds as if Synta dispensed with that part of the procedure.

 

 

 

I believe that they still try to optimize the system by centering the corrector plate.  As I have pointed out in other threads, the system is very insensitive to small decenter errors (~ < +/- 3 mm) so I believe that this step is a waste of time.  I don't know what they used to do in terms of clock angle adjustment on the corrector assembly but if they aren't currently doing it, this data suggests that they should be.

 

John

 

Back in the day "clock angle," rotation, was part of the test setup I am talking about. ;)



#32 TG

TG

    Vanguard

  • *****
  • Posts: 2357
  • Joined: 02 Nov 2006
  • Loc: Latitude 47

Posted 04 May 2017 - 04:49 PM

A few years back, the secondary on my C9.25 rotated as I was adjusting collimation. Going by the position of the collimation screws, I knew there were just three positions to try to return it to the factory position. So I tried all three and Roddier tested it at each position after collimating it. The factory position was clearly better than the other two.

 

I did not rotate the corrector as John as done but given my experience, at least for my scope, either the secondary wasn't rotationally symmetric or what Roddier was showing me was incorrect. 

 

Tanveer.



#33 Gil V

Gil V

    Apollo

  • -----
  • Posts: 1367
  • Joined: 09 Sep 2012

Posted 04 May 2017 - 09:37 PM

You are describing the exact assembly process we used at Criterion.

#34 gnowellsct

gnowellsct

    Hubble

  • *****
  • Posts: 15518
  • Joined: 24 Jun 2009

Posted 22 May 2017 - 12:09 PM

John this is an important post- script and I think you should submit it as a short second article to CN.  Otherwise we're never going to find it.  If you want you can put it up on Astromart they have a much faster review and post-policy (as in about 24 hours).  Or post it here.  It should just be available.  Greg N



#35 BillP

BillP

    Hubble

  • *****
  • Posts: 18642
  • Joined: 26 Nov 2006
  • Loc: Spotsylvania, VA

Posted 27 May 2017 - 11:08 AM

John this is an important post- script and I think you should submit it as a short second article to CN.  Otherwise we're never going to find it.  If you want you can put it up on Astromart they have a much faster review and post-policy (as in about 24 hours).  Or post it here.  It should just be available.  Greg N

 

+1 waytogo.gif



#36 jhayes_tucson

jhayes_tucson

    Fly Me to the Moon

  • *****
  • topic starter
  • Posts: 7099
  • Joined: 26 Aug 2012
  • Loc: Bend, OR

Posted 27 May 2017 - 12:20 PM

Well, I'm about to try this same measurement and alignment process on a third C14 so I may consider turning it into a more formal article if it works out.

 

John


  • Daniel Mounsey likes this

#37 spokeshave

spokeshave

    Vanguard

  • *****
  • Posts: 2033
  • Joined: 08 Apr 2015

Posted 11 March 2018 - 10:31 AM

Have you guys seen this? 
 
https://youtu.be/2TXIr-x-p1o


Reviving and old thread... does anyone know what software is used in the video for the 3D star profile?

Thanks.

Tim

#38 jsowens

jsowens

    Vostok 1

  • -----
  • Posts: 134
  • Joined: 20 Feb 2014

Posted 23 March 2018 - 07:52 PM

I too am very interested in the software used here



#39 Daniel Mounsey

Daniel Mounsey

    Vendor (Woodland Hills)

  • *****
  • Vendors
  • Posts: 7523
  • Joined: 12 Jun 2002

Posted 24 March 2018 - 12:53 AM

This is fantastic work John. 



#40 davidc135

davidc135

    Viking 1

  • -----
  • Posts: 697
  • Joined: 28 May 2014
  • Loc: Wales, UK

Posted 24 March 2018 - 03:46 AM

I believe that they still try to optimize the system by centering the corrector plate.  As I have pointed out in other threads, the system is very insensitive to small decenter errors (~ < +/- 3 mm) so I believe that this step is a waste of time.  I don't know what they used to do in terms of clock angle adjustment on the corrector assembly but if they aren't currently doing it, this data suggests that they should be.

 

John

Would this insensitivity to centering error be dependant on whether the secondary is aspherised? There's been some debate on this regarding 8 inch scts. I don't know whether they are or not but the three possibilities are: completely spherical, aspherised only to improve the wavefront or fully aspherised to eliminate coma.

Another question if I may go slightly o.t. Does higher order S.A show as zones in a null test? Would they be significant in practise in smaller instruments? I expect the zones I see on the 4inch c. plate I'm working on are due to poor technique, but I wondered. Thanks.

 

David



#41 jhayes_tucson

jhayes_tucson

    Fly Me to the Moon

  • *****
  • topic starter
  • Posts: 7099
  • Joined: 26 Aug 2012
  • Loc: Bend, OR

Posted 24 March 2018 - 10:43 AM

In the Celestron design, the secondary is nominally spherical so from a design stand point, the centering tolerance is pretty low for the whole corrector/secondary assembly.  In manufacturing, they apply minor zonal corrections to the secondary to correct for errors in the corrector.  These corrections are intended to be small amplitude and "smooth" so they shouldn't have a large impact on the centering tolerance.  If the corrections are large with steep slopes, it creates other problems down the line when they try to get the system aligned and that's an undesirable result.  My impression is that they also try to make the zonal corrections radially symmetric--even though they mark the orientation of each component.  As I've said, they don't really have a sensitive enough test to optimize the clock angle alignment between the three components.  In any manufacturing operation, the ultimate goal is to eliminate this kind of operation in the first place so I'm sure that they try hard to make everything as radially symmetric as possible.  The bottom line is that the centering tolerance of the whole corrector/secondary tolerance will generally not be very tight--even in the real world.  The assembly should be centered as well as possible but you won't see any dramatic change in performance if it is off by a "little bit" (meaning up to about 3 mm on a C14 Edge.)   The tightest tolerance is on the alignment of the secondary mirror.  It only takes very small errors to cause significant changes in performance and that's where to put most of your efforts to improve image quality.

 

As for high order SA:  The only way that I can answer that question is to tell you that 5th order SA will produce 6th order variations in the wavefront (i.e. W060*r^6).  That's a zonal variation but the slopes are well controlled.  If you null test the system and it is well corrected, you will not see any zones in the wavefront (or the slopes.)  If you see zones in a null test, it is because the system is not well corrected.

 

John

 

John


  • SleepIsWrong and davidc135 like this

#42 davidc135

davidc135

    Viking 1

  • -----
  • Posts: 697
  • Joined: 28 May 2014
  • Loc: Wales, UK

Posted 24 March 2018 - 11:00 AM

Thanks again. In figuring the corrector(s) I was pleased if there appeared a smooth 'Schmidt type' curve with no thoughts of higher order sa being corrected. So has may well contribute to the zones I see. In practise they can hopefully be removed by local figuring guided

by the knife edge null.

What is WO in the 6th order equation?

 

David


Edited by davidc135, 24 March 2018 - 11:28 AM.


#43 orangeusa

orangeusa

    Mariner 2

  • *****
  • Posts: 228
  • Joined: 25 Jul 2018
  • Loc: Orange, CA

Posted 16 August 2019 - 08:01 PM

Reviving and old thread... does anyone know what software is used in the video for the 3D star profile?

Thanks.

Tim

Yearly bump. Still wondering.

 

Just found these "rotate your corrector plate for fun and profit" threads.  lol

 

If we had a metric (other than a star test), would be nice to check your SCT alignment......and see if rotation can get you improvement for no $$$....

 

.



#44 Benni123456

Benni123456

    Vostok 1

  • -----
  • Posts: 194
  • Joined: 20 Apr 2017
  • Loc: Munich, Germany

Posted 18 August 2019 - 07:48 PM

what this means is that before buying an edge hd,

One should ask the dealer if he has an optical bench and can make some testing.

 

I guess this test could also be done with a rhonchi eyepiec, isnt it?

 

In any case it fits the overall picture:

 

Mirrors and correctors with haze inside,

Reducers that are decentered and need different collimation,

Spherochromatism that is visible with a usual cmos chip.

 

A misaligned corrector is then just something that we can add to this list.

 

If it is like that, why do they not just sell the parts and leave the assembly to the buyer?

 

Whats the point in paying for a telescope if you have to open it, clean the haze, align the corrector to somewhere else and recollimate everything if you screw on a reducer? (or even design an own reducer because the one of Celestron does not work)

 

They clearly have a severe quality control problem there.


Edited by Benni123456, 18 August 2019 - 07:56 PM.


#45 Benni123456

Benni123456

    Vostok 1

  • -----
  • Posts: 194
  • Joined: 20 Apr 2017
  • Loc: Munich, Germany

Posted 18 August 2019 - 07:55 PM

In the Celestron design, the secondary is nominally spherical so from a design stand point, the centering tolerance is pretty low for

I guess the corrector lens in the edge design complicates things.

I think it is fair to say that the edges are similarly (non-) tolerant as an rc.

Especially with the reducer. Maybe without reducer its insensitive to decentering errors. With reducer, I can not really say that.


Edited by Benni123456, 18 August 2019 - 08:00 PM.


#46 jhayes_tucson

jhayes_tucson

    Fly Me to the Moon

  • *****
  • topic starter
  • Posts: 7099
  • Joined: 26 Aug 2012
  • Loc: Bend, OR

Posted 19 August 2019 - 12:43 AM

what this means is that before buying an edge hd,

One should ask the dealer if he has an optical bench and can make some testing.

 

I guess this test could also be done with a rhonchi eyepiec, isnt it?

 

In any case it fits the overall picture:

 

Mirrors and correctors with haze inside,

Reducers that are decentered and need different collimation,

Spherochromatism that is visible with a usual cmos chip.

 

A misaligned corrector is then just something that we can add to this list.

 

If it is like that, why do they not just sell the parts and leave the assembly to the buyer?

 

Whats the point in paying for a telescope if you have to open it, clean the haze, align the corrector to somewhere else and recollimate everything if you screw on a reducer? (or even design an own reducer because the one of Celestron does not work)

 

They clearly have a severe quality control problem there.

 

It doesn't sound like you've spent a lot of time doing high-end optical testing so let me explain a few things:

 

1)  No dealer has an optical bench and very few would have the expertise to use it even if they had the right equipment.

 

2)  If you read my original report, you saw that I used about $300k worth of precision optical equipment to make these measurements.  Achieving the kind of precision that I reported is not trivial and I only estimated an absolute accuracy on the order of around 1/10 wave.  Each measurement required the better part of a day and that doesn't include the setup or break down time.  Heck, Celestron can't even make the kind of measurements that I demonstrated in these reports--and they have more equipment than any dealer!

 

3)  The Ronchi test is a slope test and while it can provide valuable information about the shape of a wavefront, it would be very difficult to use a "garden variety" Ronchi shop test to perform this kind of optical alignment.  That's why Celestron has so much trouble with this type of procedure in the first place.  This procedure requires a quantitative measurement with a very high degree of A) sensitivity to the systematic wavefront errors (i.e. not the system wavefront errors plus errors introduced by turbulence) and B) a high degree of repeatability.  Celestron uses a number of widely publicized methods including star testing and Rochi testing to align components but since these are all qualitative tests, the results are very hit or miss.

 

4)  Finally, I agree that if you want a better telescope, Celestron might not be the first choice; but, don't expect to pay Celestron prices.  Celestron can deliver a pretty high quality telescope at a pretty reasonable price, but as you've pointed out, there can be problems.  You'll pay almost double for a PlaneWave instrument for any given aperture, but you'll also get a higher quality system.  And if PW isn't good enough, you can do what the professionals do:  Custom order a system.  There are a number of vendors that supply custom astronomical equipment up to just about any size or configuration that you want.  Just be prepared to pay 10x- 40x what a Planewave system might cost for any given aperture.  These folks make a nice 0.5 m F/2 RASA-like system:  http://www.dfmengine...e_assembly.html.  Just don't even think about asking about a price.  You know what they say:  If you have to ask...you probably can't afford it.  

 

Most folks, however, can't afford to take that route and are willing to take their chances with a Celestron.  My C14 Edge has wavefront performance that will match any Planewave system.  So, it's possible to get a pretty good system from Celestron.  I've sure seen a lot of great images take with Celestrons as well so I'm not the only one who has experienced great imaging with a Celestron.

 

 

John


  • SandyHouTex likes this

#47 Benni123456

Benni123456

    Vostok 1

  • -----
  • Posts: 194
  • Joined: 20 Apr 2017
  • Loc: Munich, Germany

Posted 19 August 2019 - 06:03 AM

I know two dealers who have optical benches.

 

For  example this one here makes a ronchi test and says he takes a week before delivery to adjust the telecsopes

https://www.teleskop...listen.de/shop/

 

here you have a few test results of them

https://www.teleskop...s:::73_119.html

 

why should it not be possible for them to rotate the corrector and make a photograph of the ronchi lines every 22.5 degrees?

 

i found this only after i had bought my scope somewhere else, sadly..

 

also, when you pay a bit more (300 euroa or so), i guess even ts would deliver you what they call an interferometric test report. At least they used to.


Edited by Benni123456, 19 August 2019 - 06:15 AM.


#48 Benni123456

Benni123456

    Vostok 1

  • -----
  • Posts: 194
  • Joined: 20 Apr 2017
  • Loc: Munich, Germany

Posted 19 August 2019 - 06:37 AM

 

Celestron uses a number of widely publicized methods including star testing and Rochi testing to align components but since these are all qualitative tests, the results are very hit or miss.

My new edge hd 800 came in the following condition

 

1) Haze on the corrector, primary and secondary, showing a visible hand print.

2) the tube had strange crystals that seemed to glue on the inner tobe wall on it all over the place

3) when i put a reducer on it, it needs a different centering.

 

This suggests (especially the crystals that were adherent on the inner tube wall and the haze on every component) that they do not even look into the scope with an eye inspection when they assemble them.

 

They also do probably not even a flashlight test. How else would a telescope with hand prints on the primary and fog and particles on the tube wall leave the factory.

 

This looks like assembling by completely untrained persons who do not care.


Edited by Benni123456, 19 August 2019 - 06:43 AM.


#49 Benni123456

Benni123456

    Vostok 1

  • -----
  • Posts: 194
  • Joined: 20 Apr 2017
  • Loc: Munich, Germany

Posted 19 August 2019 - 06:48 AM

here is a better description what this dealer says he tests before delivery on his optical bench

 

https://www.teleskop...ut-us:::12.html

 

should have bought there before...

they use a 1/25 flat with a 254 lpi ronchi.

 

you used an 1/20 flat...

 

I dont know whether this would give you an edge in estimating the corrector angle...

 

 

 


Edited by Benni123456, 19 August 2019 - 07:02 AM.


#50 Richard Whalen

Richard Whalen

    Mercury-Atlas

  • *****
  • Posts: 2842
  • Joined: 24 Sep 2007
  • Loc: Florida

Posted 19 August 2019 - 10:47 AM

John,

 

Really good info, thanks for the report! Would love to see a picture of your test bench in use. Was wondering if you know the I.D. of the baffle tube on a C14?

 

Good luck with your third C14, eagerly awaiting your results.




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