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

why aren't porro prisms more often cemented?

  • Please log in to reply
80 replies to this topic

#51 Jon Isaacs

Jon Isaacs

    ISS

  • *****
  • Posts: 77770
  • Joined: 16 Jun 2004
  • Loc: San Diego and Boulevard, CA

Posted 28 May 2016 - 01:34 AM

 

 

Simple things like temperature becomes important.  That aluminum prism cage will shrink by half the total tolerance (0.004") with a mere 10 degree temperature change, whereas the glass will shrink by only 50 millionths of an inch in the same temp delta.

 

 

The CTE of aluminum is 12.5 x 10-6 /F.  A 10 degree F temperature change would result in a 0.00012" change in a 1 inch part.

 

:shrug:

 

Jon

 

 

Thanks for correction.  That was a double error -- typo, meant to type .0004; and I doubled the radius.  The number I was after was 0.00025 (10° F change over two inches)

 

But my point wasn't in the exact values, it was just that manufacturing gets much more complicated and expensive as the tolerances get smaller.     

 

 

A couple of 1/10ths is definitely not so easy, 0.004",  that's easy..An adjustment that can handle a couple of 1/10s, +/- 0.0001", that's not so easy either.

 

Jon



#52 marktd

marktd

    Mariner 2

  • -----
  • topic starter
  • Posts: 217
  • Joined: 30 Sep 2015
  • Loc: Bay Area, CA

Posted 28 May 2016 - 01:35 AM

Glenn, you refer to "perceived angular offset", as the only thing the eye cares about.  But I'm not sure how the eye directly perceives angular offset at all.

 

Isn't it really the case that the eyes perceive two images that don't perfectly align on their respective retinas, and they are actually perceiving a lateral offset of images, in the amount of the angular offset times the focal length.

 

The lateral offset increases with magnification, and thus becomes more apparent the higher the mag, like in the attached drawing.

 

 

Attached Thumbnails

  • Screen Shot 2016-05-27 at 11.24.23 PM.png


#53 marktd

marktd

    Mariner 2

  • -----
  • topic starter
  • Posts: 217
  • Joined: 30 Sep 2015
  • Loc: Bay Area, CA

Posted 28 May 2016 - 02:01 AM

 

A couple of 1/10ths is definitely not so easy, 0.004",  that's easy..An adjustment that can handle a couple of 1/10s, +/- 0.0001", that's not so easy either.

 

Jon

 

 

Completely agree. In fact, in my prior post I referred to exactly such an adjuster as "unrealistic" (though I meant in a practical sense, not mechanically -- mechanically it's not terribly hard to make such an adjuster, but in practical economic terms at bino price point, not as easy) 

 

That said, if one truly needed tenths precision, I think it's probably more practical to do it with an adjuster than with machining to that tolerance, which was my original point.



#54 Jon Isaacs

Jon Isaacs

    ISS

  • *****
  • Posts: 77770
  • Joined: 16 Jun 2004
  • Loc: San Diego and Boulevard, CA

Posted 28 May 2016 - 07:26 AM

 

 

A couple of 1/10ths is definitely not so easy, 0.004",  that's easy..An adjustment that can handle a couple of 1/10s, +/- 0.0001", that's not so easy either.

 

Jon

 

 

Completely agree. In fact, in my prior post I referred to exactly such an adjuster as "unrealistic" (though I meant in a practical sense, not mechanically -- mechanically it's not terribly hard to make such an adjuster, but in practical economic terms at bino price point, not as easy) 

 

That said, if one truly needed tenths precision, I think it's probably more practical to do it with an adjuster than with machining to that tolerance, which was my original point.

 

 

The adjuster, if you can actually design it, is probably cheaper. But when you machine it, it's a done deal and likely more reliable.

 

Jon



#55 Rich V.

Rich V.

    Fly Me to the Moon

  • *****
  • Posts: 6430
  • Joined: 02 Jan 2005
  • Loc: Lake Tahoe area, Nevada

Posted 28 May 2016 - 09:46 AM

Glenn, here's another post supporting alignment tolerances scaling inversely with magnification.  When we look through a misaligned binocular we see the apparent error which is the angular error times the magnification.

 

EdZ's full post about collimation tolerances is here:   http://www.cloudynig...wanted/?p=68300

 

He states:

 

"Generally there are three collimation errors to be concerned about. For all misalignments the allowable error is smaller as magnification increases. Actually this results in the same apparent visual error, the angular error multiplied by the magnification. My actual experience is that my eyes will not allow separation as large as defined by these limits, especially at magnifications of 15x and higher.

 

Vertical Alignment, a serious error, is when one image is higher than the other image. The eyes have no muscles to accommodate for vertical error. The allowable divergence is only 4 arcmin at 7x to 10x, only 3 arcmin at 12x to 15x and 2 arcminutes at 15x to 20x. Personally, I cannot tollerate even 2 to 3 arcminutes of vertical error even at magnifications of 10x and only 1 arcmin at 16x.

 

Horizontal Convergence, the least problematic error, is when the image in the right eyepiece is to the left of the image in the left eyepiece. You might think of this as cross-eyed. The most separation can be tolerated in horizontal convergence, 10 arcmin at 7x to 10x, 8 arcmin at 12x to 15x and 6 arcmin at 15x to 20x. Most eyes can readily accommodate this error, but these limits I find a little too wide. I could not tollerate even half that much.

 

Horizontal Divergence, another serious error, is when the image in the right eyepiece is to the right of the image in the left eyepiece. The images are spread apart, an error that the eyes cannot accommodate without strain. The allowable error for horizontal divergence is 6 arcmin at 7x to 10x and only 4 arcmin at 12x to 15x and 3 arcmin at 15x to 20x.

 

Keep in mind, the magnification of your binocular has a significant influence on how you see this error. Your eyes see the apparent error, or the angular error times the magnification. For instance a 3 arcminute error in a 10x binocular will appear to have a magnified apparent size of 30 arcminutes. That same error in a 15x binocular will have an apparent size of 45 arcminutes and in a 20x binocular will appear as 60 arcminutes.

 

For reference 1 inch at 100 yards is just slightly less than 1 arcminute."

 

Rich



#56 marktd

marktd

    Mariner 2

  • -----
  • topic starter
  • Posts: 217
  • Joined: 30 Sep 2015
  • Loc: Bay Area, CA

Posted 28 May 2016 - 10:36 AM

 

The adjuster, if you can actually design it, is probably cheaper. But when you machine it, it's a done deal and likely more reliable.

 

Jon

 

 

Definitely.

 

And of course, the third option is to rely neither on precise machining nor adjustment, but rather optical alignment during assembly and then epoxy in place.  I see this in some, but not a majority, of the binos I've disassembled, but it seems the norm in microscope components.  Is it possible that this is because the latter require higher precision?  (btw, I know I'm comparing apples to oranges in many respects, for example, binos and microscopes have very different field conditions, and I'm comparing cheap binos to high-end microscopes, simply because that's what I happened to have disassembled -- I don't have any high end binos I'm willing to disassemble and I don't have any low-end microscopes) 

 

This all relates back to my original issue -- ie, cemented prisms don't require precise machining nor adjustment.  Just align and cement once.  Seems like a good approach in many situations.  I understand that the align/cement operation may be expensive relative to cheap binos, but the cost/benefit ratio seems good at the mid/high end, yet it is so rare.   That seemed not to make sense, so I wondered if there was perhaps an optical reason that I didn't understand. 



#57 FrankL

FrankL

    Viking 1

  • -----
  • Posts: 558
  • Joined: 30 Jul 2009
  • Loc: Winnipeg, Manitoba Canada

Posted 28 May 2016 - 11:00 AM


Horizontal Convergence, the least problematic error, is when the image in the right eyepiece is to the left of the image in the left eyepiece. You might think of this as cross-eyed. The most separation can be tolerated in horizontal convergence, 10 arcmin at 7x to 10x, 8 arcmin at 12x to 15x and 6 arcmin at 15x to 20x. Most eyes can readily accommodate this error, but these limits I find a little too wide. I could not tollerate even half that much.

 

Horizontal Divergence, another serious error, is when the image in the right eyepiece is to the right of the image in the left eyepiece. The images are spread apart, an error that the eyes cannot accommodate without strain. The allowable error for horizontal divergence is 6 arcmin at 7x to 10x and only 4 arcmin at 12x to 15x and 3 arcmin at 15x to 20x.

 

This is confusing. It's my understanding that horizontal convergence is when the images are too close together and not easily accommodated by the human eye because it has difficulty diverging images whereas horizontal divergence is when the images are too far apart and is more easily accommodated by eye because it can more readily converge images.  For these reasons, horizontal divergence tolerances should be greater than horizontal convergence tolerances. 


Edited by FrankL, 28 May 2016 - 11:50 AM.


#58 Jon Isaacs

Jon Isaacs

    ISS

  • *****
  • Posts: 77770
  • Joined: 16 Jun 2004
  • Loc: San Diego and Boulevard, CA

Posted 28 May 2016 - 11:10 AM

Rich:

 

I have only superficiality been able to follow the alignment tolerances part of this thread but it seems to me that its a matter of point of view.  If you are measuring while looking though the eyepieces, then there is only one spec.  If you are thinking about the alignment of the objectives and prisms relative to the object plane, then the magnification must be consider because it affects what the eye actually sees.

 

Mark:

 

The biggest differences between the binoculars I have at home and the microscopes I use in the lab are the lives they lead.  

The binoculars are in their case, they get tossed into a bag with out stuff, put in the trunk or back of a pickup and get bounced around. In use they get bumped and sometimes dropped. The go from near freezing to 100 degree plus temperatures..  It's a tough life.

 

The microscopes are installed by a technician who aligns them.  They live in a temperature and humidity controlled environment on a rock solid bench, they are serviced regularly and if they are moved, it might happen every 10 years or so, it is done with great care.

 

It's just a different world..

 

Jon



#59 marktd

marktd

    Mariner 2

  • -----
  • topic starter
  • Posts: 217
  • Joined: 30 Sep 2015
  • Loc: Bay Area, CA

Posted 28 May 2016 - 12:05 PM

John, that's what I meant when I referred to different "field conditions", which clearly drive some of the design.  In a bino, if you drop and dent it, even permanently shifting the position of the adjuster itself, you can still adjust it back to optical correctness.  Microscopes have little need to accommodate damage, and subsequent re-alignment, so more things are aligned once and glued into place.  You drop it, you just buy a new one.

 

There are other big differences too.  $10,000 microscopes are a commodity, probably 1,000 times more sold every year than $10,000 binos, so the manufacturing economies are totally different, and this also drives design decisions.  



#60 Rich V.

Rich V.

    Fly Me to the Moon

  • *****
  • Posts: 6430
  • Joined: 02 Jan 2005
  • Loc: Lake Tahoe area, Nevada

Posted 28 May 2016 - 12:12 PM

Rich:

 

I have only superficiality been able to follow the alignment tolerances part of this thread but it seems to me that its a matter of point of view.  If you are measuring while looking though the eyepieces, then there is only one spec.  If you are thinking about the alignment of the objectives and prisms relative to the object plane, then the magnification must be consider because it affects what the eye actually sees.

 

 

 

Yes, Jon, I've come to realize that Glenn and I talking about the same thing but looking from opposite directions.

 

Many published collimation tolerances are not expressed relative to the image space so they don't reflect the apparent effect at the eyepiece.  Tolerances are useless if we don't understand what they really represent.

 

I was trying to point out that Claudio's revised tolerances (actual axial alignment tolerances x magnification) are a better way of referring to the degree of a binocular's apparent misalignment than listing tolerances the more common way as 2' allowable step error between axes, for instance (this would only apply to a 7-8x bino).  This could lead to confusion as it may be inferred that the 2' tolerance applies to a 7x bino or a 70x bino.  Claudio's approach, listing a 14' absolute apparent error is simpler (2' error @ 7x = 14' apparent/ 1' error @ 14x = 14' apparent/ .2' error @ 70x = 14' apparent, etc). I understand now that Glenn is talking about the same thing; he was just using small numbers (actual deviation tolerances between axes?) that didn't reflect the effect of magnification and it threw me off at the time because the numbers were so small.

 

Rich


  • Jon Isaacs likes this

#61 Rich V.

Rich V.

    Fly Me to the Moon

  • *****
  • Posts: 6430
  • Joined: 02 Jan 2005
  • Loc: Lake Tahoe area, Nevada

Posted 28 May 2016 - 12:19 PM

 


Horizontal Convergence, the least problematic error, is when the image in the right eyepiece is to the left of the image in the left eyepiece. You might think of this as cross-eyed. The most separation can be tolerated in horizontal convergence, 10 arcmin at 7x to 10x, 8 arcmin at 12x to 15x and 6 arcmin at 15x to 20x. Most eyes can readily accommodate this error, but these limits I find a little too wide. I could not tollerate even half that much.

 

Horizontal Divergence, another serious error, is when the image in the right eyepiece is to the right of the image in the left eyepiece. The images are spread apart, an error that the eyes cannot accommodate without strain. The allowable error for horizontal divergence is 6 arcmin at 7x to 10x and only 4 arcmin at 12x to 15x and 3 arcmin at 15x to 20x.

 

This is confusing. It's my understanding that horizontal convergence is when the images are too close together and not easily accommodated by the human eye because it has difficulty diverging images whereas horizontal divergence is when the images are too far apart and is more easily accommodated by eye because it can more readily converge images.  For these reasons, horizontal divergence tolerances should be greater than horizontal convergence tolerances. 

 

 

Hi, Frank, it is confusing;  I think this is a good explanation of this seeming contradiction:

 

http://www.cloudynig...ested/?p=323852

 

Rich


  • FrankL likes this

#62 marktd

marktd

    Mariner 2

  • -----
  • topic starter
  • Posts: 217
  • Joined: 30 Sep 2015
  • Loc: Bay Area, CA

Posted 28 May 2016 - 12:21 PM

 


Horizontal Convergence, the least problematic error, is when the image in the right eyepiece is to the left of the image in the left eyepiece. You might think of this as cross-eyed. The most separation can be tolerated in horizontal convergence, 10 arcmin at 7x to 10x, 8 arcmin at 12x to 15x and 6 arcmin at 15x to 20x. Most eyes can readily accommodate this error, but these limits I find a little too wide. I could not tollerate even half that much.

 

Horizontal Divergence, another serious error, is when the image in the right eyepiece is to the right of the image in the left eyepiece. The images are spread apart, an error that the eyes cannot accommodate without strain. The allowable error for horizontal divergence is 6 arcmin at 7x to 10x and only 4 arcmin at 12x to 15x and 3 arcmin at 15x to 20x.

 

This is confusing. It's my understanding that horizontal convergence is when the images are too close together and not easily accommodated by the human eye because it has difficulty diverging images whereas horizontal divergence is when the images are too far apart and is more easily accommodated by eye because it can more readily converge images.  For these reasons, horizontal divergence tolerances should be greater than horizontal convergence tolerances. 

 

 

I think it's mistaken to think of the eyes as "diverging images".  Rather the eyes converge onto the two images that are slightly closer together than they should be, and the eyes can merge these into one image.  The eyes do this naturally (have the supporting muscles), when focusing from near to infinity.  If the images are further apart (diverged) than they should be, the eyes cannot diverge (angle outward beyond parallel) to then merge the images.


  • FrankL likes this

#63 GlennLeDrew

GlennLeDrew

    Hubble

  • *****
  • Posts: 15849
  • Joined: 17 Jun 2008
  • Loc: Ottawa, Ontario, Canada

Posted 28 May 2016 - 12:27 PM

Rich,

[Composed while the previous three posts appeared...]

 

OK, I see what Claudio means, and his 'absolute' angle tolerances are in line with my own thinking. Stating that magnification is accounted for is unfortunate, for this can so easily be misconstrued. It's intended to mean *after* the image has been magnified, not before. In other words, every image, at whatever magnification, is to be treated to a common standard, as like a unit magnification view seen by eye alone. In science this is called normalization.

 

If one is looking through a battery of binos of unknown magnification at a single star, the permissible image offset as perceived by the eyes is unchanged for all. If it's X arcminutes at 2X, it's X arcminutes at 20X.

 

Now, if you're looking at a double star and are using this as a measuring reference, then magnification plays a role of course. This is what it appears EdZ has done. And I can see why one might take such a route if no apparatus is to hand whereby the angular offset behind the eyepieces can be measured directly.

 

The process of collimation typically involves the superposition of one (magnified) eyepiece image against a view of the target seen directly by the comparator. The latter image need not be magnified at all. Such a scheme directly measures the error as perceived by the eyes, which is independent of the instrument's magnification. Whether the bino delivers 2X or 20X, the operator detects offset to the same 'absolute' limit, and can adjust to only the same limit. That limit being determined by his eye's resolving power, or about 1'.

 

If the comparator itself magnifies, the operator's ability to detect and adjust to better precision scales accordingly. But if tolerances allow offsets larger than the eye's resolving power, a magnifying comparator is unnecessary. And in the context of an operation eyeing the bottom line, extra time taken to adjust to better-than-good alignment is unlikely to be benevolently sanctioned.


Edited by GlennLeDrew, 28 May 2016 - 12:29 PM.

  • Rich V. likes this

#64 GlennLeDrew

GlennLeDrew

    Hubble

  • *****
  • Posts: 15849
  • Joined: 17 Jun 2008
  • Loc: Ottawa, Ontario, Canada

Posted 28 May 2016 - 12:35 PM

Nowadays factory collimation involves sophisticated equipment whereby the operator watches a screen and tweaks until a dot is made to move to within the required distance of a cross hair. But the same optical limits apply, and I suspect the same apparatus is used unchanged for all models of bino, irrespective of their magnification.



#65 FrankL

FrankL

    Viking 1

  • -----
  • Posts: 558
  • Joined: 30 Jul 2009
  • Loc: Winnipeg, Manitoba Canada

Posted 28 May 2016 - 12:37 PM

"This all relates back to my original issue -- ie, cemented prisms don't require precise machining nor adjustment.  Just align and cement once.  Seems like a good approach in many situations.  I understand that the align/cement operation may be expensive relative to cheap binos, but the cost/benefit ratio seems good at the mid/high end, yet it is so rare.   That seemed not to make sense, so I wondered if there was perhaps an optical reason that I didn't understand."

 

Optically, it seems that cemented Porro I prisms have been found to perform well in relatively low powered hand-held binoculars as evidenced by the Bell & Howell M19 7x50, Kern/Leica 8x30, Avimo 7x42, and a 1980's line of Swarovski Habicht SL's.  However, historically Porro II prisms were cemented instead of Porro I's because as noted in previous posts it was a lot easier to construct a prism cage supporting Porro II prisms than supporting Porro I's. Regardless, during WW II many Porro II arrangements had high occurrences of prism separation and chipping. Later, by the 1970's it became feasible to cement Porro I's because the UV cure cement which was used had much greater adhesion strength than the Canada balsam used during WW II. Going into the 1990's with multi AR coatings now capable of 90%+ light transmission, cementing prisms to increase light transmission just a small percent became superfluous.


Edited by FrankL, 28 May 2016 - 01:08 PM.


#66 marktd

marktd

    Mariner 2

  • -----
  • topic starter
  • Posts: 217
  • Joined: 30 Sep 2015
  • Loc: Bay Area, CA

Posted 28 May 2016 - 02:01 PM

Frank, do you think balsam-cemented Porro II prism separation rates were higher than lens (objective or eyepiece) separation rates, and if so, any theory why?  Maybe due to the higher and more frequent forces applied to prisms due to cantilevering, etc.  Maybe thermal induced stresses, which perhaps apply more unequal forces on prisms than on lenses?  



#67 FrankL

FrankL

    Viking 1

  • -----
  • Posts: 558
  • Joined: 30 Jul 2009
  • Loc: Winnipeg, Manitoba Canada

Posted 28 May 2016 - 02:24 PM

"Frank, do you think balsam-cemented Porro II prism separation rates were higher than lens (objective or eyepiece) separation rates,"  - Absolutely.

 

"and if so, any theory why? Maybe due to the higher and more frequent forces applied to prisms due to cantilevering, etc.  Maybe thermal induced stresses, which perhaps apply more unequal forces on prisms than on lenses?" - The reasons are varying degrees of cantilevering forces on the cemented prisms, thermal stress and balsam quality.

 

Depending on the build of the prism cage some WW II Porro II binoculars seem to have had higher incidences of prism separation than others f.ex. the Barr & Stroud Porro II models had a lot of problems and Zeiss not nearly as many. Leitz Porro II separation rates seem lower than the Barr & Strouds, but the Leitz prisms seemed highly prone to chipping possibly due to expansion/contraction of some of the components in the prism cage caused by changes in temperature. 


Edited by FrankL, 28 May 2016 - 02:27 PM.


#68 Rich V.

Rich V.

    Fly Me to the Moon

  • *****
  • Posts: 6430
  • Joined: 02 Jan 2005
  • Loc: Lake Tahoe area, Nevada

Posted 28 May 2016 - 03:25 PM

Nowadays factory collimation involves sophisticated equipment whereby the operator watches a screen and tweaks until a dot is made to move to within the required distance of a cross hair. But the same optical limits apply, and I suspect the same apparatus is used unchanged for all models of bino, irrespective of their magnification.

 

Years ago now, I visited Bill Cook while I was up in Seattle.  He showed me the Fuji UBMM collimator and it was fascinating to see how easy it was to check collimation; similar to standing the bino onto an overhead projector.  Very cool.

 

The only collimation I ever need to do is on BTs with rhombs so it's a simple "eyeball" affair on a distant target; no tolerances or fancy equipment necessary.  No hinge axis either, technically...   ;)

 

Rich



#69 Gordon Rayner

Gordon Rayner

    Gemini

  • *****
  • Posts: 3248
  • Joined: 24 Mar 2007

Posted 28 May 2016 - 04:36 PM

   Several posts  back,  above,  
Rich V  mentioned   binoculars  with   two  hinge axes,(  hopefully   manufactured  parallel  to each other).    

 

Changing the IPD   on such   , even from 56mm to 74 mm.    will have   a much smaller  effect  upon  the parallelism , or lack of it,  of the two optical  axes,   than would   a 56mm to 74mm  swing   of a  one-hinged  handheld   which  were  to  suffer from lack of parallelism   of its  hinge and  one or both optical  axes.

 

To check  the  parallelism   of   either optical  axis  with   its hinge, a  peek- around rhomboidal  beam combiner  (qv via  searching this forum )  can be  used, at least  if the line of sight    is a straight   shot , not inclined, toward the target.   Your servant  has done this  on   Porro II's   including Nikon 20 x 120, I and II, Fuji 25 x 150  I,  and Fuji 15 x 80,  while prism sliding in tiny increments, sometimes via  fine pitch setscrews  added  to  minimize  tiny  guesswork increments,   after  setting the eccentrics to neutral ( or leaving the eccentrics  in  wherever  they  are locked by  bad corrosion) .

Inclined binocs  require  additional equipment.

 

Swing the hinge   through  a large  angle  ( with the IPD  connector  disconnected),  60,  180, or even 360  degrees  rotation.   Observe the   motion  of the  direct target view  superimposed upon the  target image seen through  that side of the binocular.   Ideally, there is no   relative motion.   

 

If there is motion,   one can use  the storied (deja vu, use search  engine)  mental  isosceles  triangle   as a straightforward  adjustment  guide  with a minimum of  steps  in a logical, vs. guesswork,    procedure.

 

In practice,  setting   2-hinge  equipment   at 65mm,  often suffices,  if  the lines of sight  are  nearly 

parallel,  under the reasonable assumption  that the hinges of   any but the most sloppy  manufacture  are parallel.  

 

Please note the absence here   of   the word  "collimation",  with   its multiple  meanings  ,   so  often confused. 

It means  , in the neighbor  ATM  forum,   something entirely different   than   often  used here.    For this  binocular forum,  what  should one  write  if describing   adjustment  of an objective lens, after passage through  reflecting prisms  or mirrors,  such that  its optical  axis  and  the optical axis of its  eyepiece   are colinear,   that is  that they  are the  same line,  not two parallel lines?   For that, collimation is the correct word.


Edited by Gordon Rayner, 28 May 2016 - 05:02 PM.


#70 Robert A.

Robert A.

    Viking 1

  • *****
  • Posts: 510
  • Joined: 21 Jan 2005
  • Loc: Johnson City TN, South Central USA

Posted 28 May 2016 - 07:07 PM

I had heard that the Docter Nobleim had cemented prisms.  From the Docter site, I believe that the "sealed with putty" is ad copy to describe the cementing process.   I went to look for the documentation. 

 

On the Docter website, it talks of putty to seal around the prisms.  "The porro prisms are on the inside sealed with putty, thus reducing the number of glass/air boundary surfaces and lowering the portion of reflected light."

       at :  http://www.docter-ge...lem-series.html

 

Dr. Holger Merlitz  says that the Docter Nobleim 8x50 B/GA is

 

"The light transmission of a binocular depends to a large extent on the quality of coating, but also on the number of optical elements to be passed, and here the Docter is certainly in a better position. Its objectives are not air spaced, the ocular construction is simpler, with less number of lens-elements, and the prisms are cemented to avoid another air-to-glass transit."

  at:  http://www.holgermer.../zeiss8x50.html

 

I would believe that FrankL would also know about this. 

 

Rob.



#71 FrankL

FrankL

    Viking 1

  • -----
  • Posts: 558
  • Joined: 30 Jul 2009
  • Loc: Winnipeg, Manitoba Canada

Posted 28 May 2016 - 07:37 PM

I would believe that FrankL would also know about this. 

 

Good observation. I missed that one. "The porro prisms are on the inside sealed with putty" doesn't sound like they're cemented between the prism faces (what is the refractive index of putty??), but, then again, Merlitz says they are cemented and he's seldom wrong. 


Edited by FrankL, 28 May 2016 - 11:20 PM.


#72 Gordon Rayner

Gordon Rayner

    Gemini

  • *****
  • Posts: 3248
  • Joined: 24 Mar 2007

Posted 29 May 2016 - 05:19 PM

Re: Parallelism of optical axes with hinge axis or axes, misnomered aka alignment or "collimation"

In # 47, there is my half-baked reference to L. C. Martin's classic work, in two volumes, Technical Optics. Vol I. is 1948, Vol. II is 1960, both Pitman, London.

They are linkable online, if one Googles " LC Martin Technical Optics, Pitman" or something similar. Volume II, p. 180 has a clear exposition of "adjustment of the axes" requirements, Fig. 134.

Martin states that eyeball convergence of 20-25 minutes of arc gives little or no discomfort.

7-8 minutes of divergence "may be managed without discomfort"

"Enforced hyperphoria (up-down) of one eye should remain below 7 or 8 minutes of arc" .

His example is optical axes mounted at an angle of alpha. Parallel rays enter the centers of the entrance pupils of a binocular instrument of magnification M. Those rays will emerge from the centers of the exit pupils at a mutual angle (M-1)*alpha ( That is, (M-1) multiplied by alpha.

Thus, if M=6, divergence of axes, alpha, should be fewer than 7.5 min/(6-1) =1.5 minutes, to avoid discomfort if the eyes must diverge.

His human tolerances imply, for the M=6x example, an alpha below 1.5 arc minutes for vertical or "step" up-down satisfaction, and alpha under 4.5 minutes to avoid discomfort if the eyes converge.

There is much more in that book(s) well worth pursuit by anybody in this forum or the neighbor ATM forum including pp. 61-63, which discusses the offset of the eyeball rotation center from the exit pupil. That causes vignetting or clipping. Large exit pupils help here, especially for wide fields scanned by the eyeball while the binocular does not move. ( WW II 10 x 80 Flaks, for example , retain their comfort for persons with decreased pupil size).

Let me see if can link it. If not successful, help will be welcome. Or, just Google.


https://archive.org/...nicalOpticsVol2

Edited by Gordon Rayner, 29 May 2016 - 05:34 PM.

  • Rich V. and patta like this

#73 Gordon Rayner

Gordon Rayner

    Gemini

  • *****
  • Posts: 3248
  • Joined: 24 Mar 2007

Posted 30 May 2016 - 01:57 PM

In # 69, above,    adjust    until   there is no   relative motion  of the two images   from the beamcombiner, which might  be a JTII  style  or a reversed, eyepiece-less   binoviewer  from a telescope( preferably with  a unilateral filter added  for   ease in  image identification),     during the  large swing.  No  isosceles  or  other  mentally visualized   triangles  are needed  here.

 

This   is    related   to  Hanna's (or Brower's  Mare Island or earlier) second  method   , in  ATM  III or   II,   for   singly hinged   handheld   binoculars.


Edited by Gordon Rayner, 30 May 2016 - 02:03 PM.


#74 patta

patta

    Explorer 1

  • -----
  • Posts: 63
  • Joined: 12 Sep 2018
  • Loc: town of Eternal Rain, Norway

Posted 08 October 2018 - 04:15 PM

Back to the origins, a little out from this thread's alignment direction...

 

On youtube, the "how it's made" video on binocular;  https://www.youtube....h?v=tKSholGNWW0

They show the cementing of some modern expensive porro I

 

- Simple mechanical jig, just a little more practical than  post #15 (no optic alignment seems)

- no baffle

- Prism mounted in the body in 2,5 seconds (plus alignment & curing)


Edited by patta, 08 October 2018 - 04:37 PM.


#75 MartinPond

MartinPond

    Aurora

  • *****
  • Posts: 4844
  • Joined: 16 Sep 2014

Posted 09 October 2018 - 12:41 AM

Cemented and 'locked' prisms often chip when dropped.

Some have a glue that gives its life, though...

Others are easy to re-align.  If their shelf was machined well

and the eccentrics were done right, the alignment just needs

debris-cleaning or popping back into the well.

 

Interesting video....the prisms are glued to each other,

and glued into the plastic holding basket. That's a lot of mass.

A bit frightening...no recovery from a drop...cracked basket.

 

"----------

Martin states that eyeball convergence of 20-25 minutes of arc gives little or no discomfort.

--------------"

 

A little under 1/2 degree... 


Edited by MartinPond, 09 October 2018 - 12:50 AM.



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