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

Converting from PV error to Seeing conditions

ATM beginner mirror making optics
  • Please log in to reply
16 replies to this topic

#1 Taosmath

Taosmath

    Viking 1

  • *****
  • topic starter
  • Posts: 945
  • Joined: 21 May 2014

Posted 17 July 2019 - 10:47 PM

Where I live seeing is rarely much better than 2 arcseconds and I have never seen it better than 1 arc second (as estimated by the doubles I can split with a well collimated scope with good optics).

 

Our club is contemplating buying a 30-32" mirror for our dob and I was asking myself the question 'how good a mirror is good enough?'.

 

To explain a little.  If we go ahead with this project, the scope would be pretty much restricted to being used within 50 miles of here, so we won't be taking it to locations with super seeing (we do have dark skies nearby though).  If the seeing is always lousy, visual images will be dominated by the seeing.  If that argument is correct, the only reason to pay for a 1/10 wavelength PV mirror is if your seeing is good enough to show extra detail when using that mirror, over a say 1/4 wavelength PV mirror.  

 

I realize that some people will always want to have the best mirror they can, so that they can take advantage of nights when seeing is excellent.  But if your seeing is NEVER better than 1 arc second, how good a mirror do I need? The scope would of course largely be used for DSO's, and if we wanted planetary we would use an aperture mask.

 

Is there some analog to the  'adding errors in quadrature' method that would let me calculate that?  If so, what is it and what is the conversion from seeing to PV error?

 

Maybe I'm thinking of this completely wrongly, in which case  would appreciate someone explaining it to me .

 

Thanks

 

Colin

 

 



#2 Oregon-raybender

Oregon-raybender

    Optical Research Engineer

  • -----
  • Vendors
  • Posts: 571
  • Joined: 13 May 2010
  • Loc: Oregon, South Western Coast

Posted 18 July 2019 - 12:31 AM

I would try for 1/8 wave (to yield 1/4 at focus) PV and 1/10 or better secondary. Smoothness (RMS) is also very important, you want at least 1/20 wave. Others may have other standards, but it's time and money that decides the limits as well. The customer has to decide what is important. You points are valid, but you still want to have a spec that will yield great views if the seeing improves.

 

Starry Nightswaytogo.gif


  • gregj888 likes this

#3 davidc135

davidc135

    Viking 1

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

Posted 18 July 2019 - 02:09 AM

That is p-v of the surface of the mirror? Or wave front?  David



#4 Vla

Vla

    Messenger

  • *****
  • Posts: 496
  • Joined: 08 Mar 2014

Posted 18 July 2019 - 08:15 AM

Well, in 1 arc sec seeing the green (~550nm) atmospheric coherence length (r0) is about 14cm, or nearly 6". It usually fluctuates within +/-50%, so the best seeing moments you can expect for a meaningful amount of time are with the coherence length of about 10". That makes the D/r0 ratio for 30-32" aperture (D) little over 3. The corresponding long-exposure RMS error is 0.162(D/r0)^(5/6), or little over 0.4 waves, with the corresponding seeing Strehl below 0.09 for long exposure, and about 0.35 for short exposure. Visual image for this D/r0 ratio is roughly midway between, or 0.20-0.25 Strehl (since the main component of atmospheric error, wavefront tilt, is still partly compensated for by eye being able to follow the image motion it causes). In terms of primary spherical aberration, this is at the 2/3 waves p-v level. Since the 1/10 wave p-v Strehl (0.96-97) is about 20% higher than 1/4 wave, your gain would have been, say, 0.18 vs. 0.22 seeing Strehl (w/o central obstruction effect), or, in terms of spherical aberration, 0.7 vs. 0.65 wave p-v. Since the Strehl figure also represents the contrast averaged over the range of MTF frequencies, the key figure is that the 1/10 mirror always have its 20% averaged contrast advantage. For some sort of visual illustration, taking defocus as the closest in effect to the seeing error, the MTF graph shows the difference between the two. Not much to talk about.

Attached Thumbnails

  • spv.png

  • Mike I. Jones, Oregon-raybender, MKV and 3 others like this

#5 MKV

MKV

    Cosmos

  • *****
  • Posts: 8087
  • Joined: 20 Jan 2011
  • Loc: Florida, USA

Posted 18 July 2019 - 10:48 AM

I would try for 1/8 wave (to yield 1/4 at focus) 

I know that you know better, and are trying to be supportive and optimistic, but the reality is stark: that's a toll order for an earth-bound telescope of that size  -- and most likely a fast focal ratio -- even if Mother She is cooperating!

 

To get 1/4 wave wavefront error at an eyepiece's last surface (exit pupil) is not the same as having an isolated smooth 1/8 ptv surface primary mirror on a workbench.  There are other contributing factors which must be accounted for in a compete telescope under actual nighttime observing conditions (tube currents, central obstruction, mirror substrate, supports, and even the eyepiece itself) that will come into play.

 

The exit pupil wavefront error is a sum-total of all image-deforming contributors in an optical train. To get a true 1/4 wave ptv wavefront at the exit pupil requires much more than an 1/8 wave ptv surface mirror.  


Edited by MKV, 18 July 2019 - 10:48 AM.

  • Oregon-raybender likes this

#6 Oregon-raybender

Oregon-raybender

    Optical Research Engineer

  • -----
  • Vendors
  • Posts: 571
  • Joined: 13 May 2010
  • Loc: Oregon, South Western Coast

Posted 18 July 2019 - 01:04 PM

Yes, I know I left out a lot of the other stuff. Gene Cross gave a great lecture on all of the above at RTMC decades ago. We are stuck with the long list of what "if's" I would add the eye to the list as well.

 

I'll suggest make it the best you can within your limits, money, people, skills and site. Put it all together and hope for the best. Group 70 is a great example of this. They had the best optical tech help, machines, testing and now building the tube and mount. The site is yet to be known.  Another example is Fremont Peak Observatory 32 inch scope (also built by Kevin) I was part of both groups (in support (I was VP) and skill (machine shop) ) Both were built by very active folks in the SF Bay area.

 

 

Starry Nightswaytogo.gif

 

https://www.fpoa.net/


Edited by Oregon-raybender, 18 July 2019 - 07:09 PM.


#7 Mike Lockwood

Mike Lockwood

    Vendor, Lockwood Custom Optics

  • *****
  • Vendors
  • Posts: 1575
  • Joined: 01 Oct 2007
  • Loc: Usually in my optical shop

Posted 18 July 2019 - 03:29 PM

One-arcsecond seeing is actually pretty good compared to many places, like the midwest.

 

However, even here in the midwest, a very good quality mirror will perform better than a poor one in steadier moments, and the difference will be noticed in those moments.  Improved contrast is also noted.  Therefore I recommend not compromising on optical quality.

 

As is alluded to by others above, it will be very important to pay attention to other factors, like equilibration, optical support, and maintaining collimation.  These typically cause more error than the optics.  Only when these factors are addressed can one achieve optic-limited performance, and many instruments never get there.

 

One question - if you have never seen better than one-arcsecond seeing, are you sure the instruments that you have made that evaluation with have addressed the factors that I mentioned above, and are not being limited by them, too?


Edited by Mike Lockwood, 18 July 2019 - 04:45 PM.

  • Dale Eason, LarsMalmgren, Oregon-raybender and 1 other like this

#8 Fatcat

Fatcat

    Vostok 1

  • *****
  • Posts: 109
  • Joined: 11 Nov 2007
  • Loc: Canada

Posted 19 July 2019 - 02:33 AM

I think this speaks directly to your question

 

http://www.whichtelescope.com/air.htm


  • gnabgib and Taosmath like this

#9 Richard Whalen

Richard Whalen

    Mercury-Atlas

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

Posted 19 July 2019 - 10:25 AM

For your location you might be better off with a 25" mirror. I have looked through dozens of premium dobs 18" to 42" and have never seen one that was close to 1/10 wave P-V. If you can get a true 1/5 or 1/6 wave P-V in the 20" to 25" range you are very lucky indeed.



#10 Taosmath

Taosmath

    Viking 1

  • *****
  • topic starter
  • Posts: 945
  • Joined: 21 May 2014

Posted 20 July 2019 - 07:36 PM

One-arcsecond seeing is actually pretty good compared to many places, like the midwest.

 

....

 

One question - if you have never seen better than one-arcsecond seeing, are you sure the instruments that you have made that evaluation with have addressed the factors that I mentioned above, and are not being limited by them, too?

Thanks for the comments.

 

To be precise, when I said that had never seen seeing better than one arc second, while that is true, I think the best i have ever seen is about 1.5 or 1.6 arc seconds.

 

And to answer your second point, no, I cannot be sure that the seeing I got was not limited by the optics I used.  Is there a way I can tell that?



#11 Taosmath

Taosmath

    Viking 1

  • *****
  • topic starter
  • Posts: 945
  • Joined: 21 May 2014

Posted 20 July 2019 - 07:37 PM

Well, in 1 arc sec seeing the green (~550nm) atmospheric coherence length (r0) is about 14cm, or nearly 6". It usually fluctuates within +/-50%, so the best seeing moments you can expect for a meaningful amount of time are with the coherence length of about 10". That makes the D/r0 ratio for 30-32" aperture (D) little over 3. The corresponding long-exposure RMS error is 0.162(D/r0)^(5/6), or little over 0.4 waves, with the corresponding seeing Strehl below 0.09 for long exposure, and about 0.35 for short exposure. Visual image for this D/r0 ratio is roughly midway between, or 0.20-0.25 Strehl (since the main component of atmospheric error, wavefront tilt, is still partly compensated for by eye being able to follow the image motion it causes). In terms of primary spherical aberration, this is at the 2/3 waves p-v level. Since the 1/10 wave p-v Strehl (0.96-97) is about 20% higher than 1/4 wave, your gain would have been, say, 0.18 vs. 0.22 seeing Strehl (w/o central obstruction effect), or, in terms of spherical aberration, 0.7 vs. 0.65 wave p-v. Since the Strehl figure also represents the contrast averaged over the range of MTF frequencies, the key figure is that the 1/10 mirror always have its 20% averaged contrast advantage. For some sort of visual illustration, taking defocus as the closest in effect to the seeing error, the MTF graph shows the difference between the two. Not much to talk about.

Vla,  thanks for this comprehensive explanation.  I'll read through it a few more times and then I may have questions



#12 Taosmath

Taosmath

    Viking 1

  • *****
  • topic starter
  • Posts: 945
  • Joined: 21 May 2014

Posted 20 July 2019 - 07:39 PM

I think this speaks directly to your question

 

http://www.whichtelescope.com/air.htm

Fatcat,

 

thanks for this.  This article, together with Vlasti post should enable me to understand this issue a little better.

 

The article was so good, I poked around more on that site.  It says it's under development.  Is that accurate or is it someones uncompleted project?  None of the links i clicked would work.



#13 Mike Lockwood

Mike Lockwood

    Vendor, Lockwood Custom Optics

  • *****
  • Vendors
  • Posts: 1575
  • Joined: 01 Oct 2007
  • Loc: Usually in my optical shop

Posted 20 July 2019 - 07:55 PM

Thanks for the comments.

To be precise, when I said that had never seen seeing better than one arc second, while that is true, I think the best i have ever seen is about 1.5 or 1.6 arc seconds.

And to answer your second point, no, I cannot be sure that the seeing I got was not limited by the optics I used.  Is there a way I can tell that?

You either need to know for sure the quality and condition of the instruments, and that they should perform well beyond the seeing, or you need to take them to a superb location and see what they can do under really good conditions.

 

Beware that article - it has a number of typos, some poor grammar, and some advice that I find quite erroneous:

 

"Advice commonly given to amateur astronomers is that it is always essential to let your telescope cool down to outdoor ambient temperature before use. This advice is misleading, as in many cases there is no advantage to be gained in letting your telescope cool. If you mostly observe deep-sky objects at low to medium magnifications, then you probably aren't being bothered much by telescope thermals."

 

That is so obviously wrong - deep sky observing yields far better detail when the telescope is cooled off.  (If you don't use high enough power, that is simply poor observing.)

 

If, however, you're a lunar and planetary observer who wants to use high magnification, then, yes, your telescope may perform better if you allow it to cool. But only if the seeing conditions at the time are very good. In average to poor seeing conditions, and especially in towns, atmospheric turbulence will degrade your views to a far great extent than telescope thermals."

 

That is also wrong - mirror cooling can have a far larger effect on views than seeing, and in my experience, with larger instruments, it is one of the largest factors in telescope performance.

 

What we're not seeing mentioned enough here is if you want to see faint things, you need big opticsIf you want to see fine detail, you need good optics and you need to make sure that equilibration, mirror mounting, and holding collimation, are all addressed.


Edited by Mike Lockwood, 21 July 2019 - 09:40 AM.

  • Asbytec likes this

#14 Asbytec

Asbytec

    Guy in a furry hat

  • *****
  • Posts: 15507
  • Joined: 08 Aug 2007
  • Loc: Pampanga, PI

Posted 20 July 2019 - 09:31 PM

A couple more statements from the site I find are both true and possibly misleading. 

 

"The implications of air turbulence is that, as you choose bigger and bigger scopes, so the optical quality becomes less of an issue."

 

This is essentially what the OP is asking. I think the jury is still deliberating, and some comments in this thread support better optics. Folks will explain that it's all about total RMS implying a high RMS rapidly drops below the diffraction limit in any seeing while a low RMS value can sustain some level of seeing and remain diffraction limited. I am not sure if seeing really operates that way, but it might. The bottom line is, the disturbed seeing disc is generally much smaller in a large aperture. At least until the larger scope becomes bloated and experiences speckling behavior. After that point, all bets are off. I understand the rule of thumb is it might reach the resolution of 1/3D or about 10" aperture. It takes some pretty bad seeing to get bloated stars and a seeing disc the size of a 4" aperture. The real bane of a larger aperture is bloated stars. 

 

"The FWHM of the seeing disc (or just Seeing) is usually measured in arcseconds, abbreviated with the symbol ("). A 1.0" seeing is a good one for average astronomical sites. So the best resoluton would be achieved with a telescope of about 4 inches aperture. Using a larger scope would not provide sharper views, except for perhaps short glimpses of greater detail."

 

I really do not understand using 1" arc second seeing to evaluate short exposure visual observing, as I understand it this is used in longer exposure imaging where the star image is 1" arc after being exposed over longer intervals. The image in a 32" is not likely going to be 1" arc in pretty good 1" arc seeing, but more likely sub arc second most of the time. So, it should blow the socks off a 4" aperture on a regular basis. Especially is, as stated, all scopes including the OP are limited to resolution of about 10" aperture due to the properties of the atmosphere at never better than 0.4" arc seeing (which is another long exposure measure). Short exposure visual snap shots will best a 10" aperture. 

 

I believe the 32" will never be diffraction limited and we'll likely never see a textbook diffraction pattern, but it's seeing induced star images can be pretty small when seeing is descent. Seeing will be what seeing will be, so it's often best to focus on things we can control such as thermal stability and attaining (and keeping) good collimation. I'd imagine in all seeing conditions one might want to keep the energy as centralized as possible to minimize the loss of contrast in all seeing conditions. I'd recommend shooting for the best optics that are feasible and affordable. At least descent.  



#15 dan chaffee

dan chaffee

    Ranger 4

  • -----
  • Posts: 312
  • Joined: 22 Mar 2006
  • Loc: North Carolina

Posted 21 July 2019 - 02:31 AM

Putting an honest 1/14th RMS,.8 strehl ratio, 1/8 wave PV..however you want to define it, on a mirror of that size

is not a cheap or easy feat.   I'm assuming this is f/4-f/5 or so.   What no one seems to point out so far is that

such an instrument could be stopped down to 11 or 12 inches of unobstructed aperture for utterly phenomenal

planetary detail on nights of 1arcsec or better seeing.   But lets consider  it at full aperture under good seeing.

At 500x it would only be 16.6 x per inch of 30 inch aperture! The shear brute force of that much light, resolving

power and contrast function at anything close to .8 strehl at the eyepiece would produce incredibly detailed images of the most demanding bright targets at that  magnification in good seeing..and well beyond in REALLY good seeing. Saturn's Enke gap should be visible at that point with it. Break out the neutral density filter for the planets.   It is extremely unlikely you would ever see a star's diffraction pattern intact with  30+ inches of aperture, but so what?   Deep sky objects would be mesmerizing with such an instrument. 


Edited by dan chaffee, 21 July 2019 - 03:03 AM.


#16 Asbytec

Asbytec

    Guy in a furry hat

  • *****
  • Posts: 15507
  • Joined: 08 Aug 2007
  • Loc: Pampanga, PI

Posted 21 July 2019 - 03:10 AM

..resolving power and contrast function at anything close to .8 strehl at the eyepiece would produce incredibly detailed images... 

 

It is extremely unlikely you would ever see a star's diffraction pattern intact with  30+ inches of aperture, but so what?   

Well said and, yes, so what.

 

One thing I keep mulling over is the idea of the Strehl performance with such a large aperture. The area of the brighter portion of the PSF is so small (what, about 0.2" arc?) so maybe Strehl might not even matter much. We're likely not going to see it, anyway, given the average (or even good) seeing. It's just a (hopefully) small, sub arc second, bright spot in the dark of space with most of it's energy centralized to the point seeing allows it to be. (The largest scope I've owned is an 18" Dob, and never even thought about it's large aperture as being a problem.) 



#17 gr5org

gr5org

    Vostok 1

  • -----
  • Posts: 107
  • Joined: 06 Sep 2018

Posted 21 July 2019 - 01:16 PM

>I really do not understand using 1" arc second seeing to evaluate short exposure visual observing,

 

I think that web page is referring to "lucky imaging" where you use a camera to take many pictures every second - thousands of pictures of the same spot in the sky - and you use software to find the 1% (or fewer) that have the best seeing.

https://en.wikipedia...i/Lucky_imaging

 

There is an observatory near me with a 24 inch (or is it 25 inch?) that uses this technique and the images are pretty stunning.


  • Asbytec likes this


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





Also tagged with one or more of these keywords: ATM, beginner, mirror making, optics



Cloudy Nights LLC
Cloudy Nights Sponsor: Astronomics