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

#1 marktd

marktd

    Mariner 2

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

Posted 24 May 2016 - 01:06 AM

I've been tinkering with frankenbinos and wondering why most porros are not cemented.  I see a few are, but the vast, vast majority are not.  There must be a good reason??


  • patta likes this

#2 GlennLeDrew  Happy Birthday!

GlennLeDrew

    Hubble

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

Posted 24 May 2016 - 03:29 AM

I'm assuming you mean the Porro I type, for the Porro II very commonly has its three parts cemented.

 

Cementing the two glass blocks would complicate their supporting, by losing the handily large, flat, hypotenuse faces. Moreover, this would eliminate the fairly important baffle aperture at the interface between the two prisms.



#3 MartinPond

MartinPond

    Aurora

  • *****
  • Posts: 4948
  • Joined: 16 Sep 2014

Posted 24 May 2016 - 05:30 AM

I think the OP is talking about having blobs of cement

   to hold the prism in place on the shelf.

 

Most of the prisms I clean are trapped between

indentations that were put in the walls of the 'shelf' the

prism is on.  Sometimes you can hear the prism

'click' back intto place.


  • patta likes this

#4 Mark9473

Mark9473

    Voyager 1

  • *****
  • Posts: 11271
  • Joined: 21 Jul 2005
  • Loc: 51°N 4°E

Posted 24 May 2016 - 07:01 AM

I don't think that's what most of us understand by cemented prisms, Martin.


  • StarDustBin likes this

#5 marktd

marktd

    Mariner 2

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

Posted 24 May 2016 - 08:52 AM

Sorry I wasn't clear.  Yes, I meant Porro I and I meant why not cement the two prism faces to each other.   

 

Glenn, does a baffle aperture between the two faces function significantly differently than at the entrance or exit? 

 

Related to that, I also notice the microscope prisms frequently have apertures painted directly on prisms, as in attached photo.   I've seen this on one binocular, but not the couple dozen others I've inspected.  Any thoughts on this practice?

 

 

Attached Thumbnails

  • prism.jpg


#6 FrankL

FrankL

    Viking 1

  • -----
  • Posts: 575
  • Joined: 30 Jul 2009
  • Loc: Winnipeg, Manitoba Canada

Posted 24 May 2016 - 09:27 AM

I'm assuming you mean the Porro I type, for the Porro II very commonly has its three parts cemented.

 

Cementing the two glass blocks would complicate their supporting, by losing the handily large, flat, hypotenuse faces. Moreover, this would eliminate the fairly important baffle aperture at the interface between the two prisms.

Am I correct that the baffle aperture is important for blocking stray light?

 

None of the cemented Porro I's I've seen (in the Avimo 7x42, Bell & Howell M19 7x50, Leica Kern AARAU 8x30) had baffles at the cemented joining of the two prisms. However, cemented prisms can be baffled. The WW II Leitz 7x50 Porro II's (and many of the Zeiss Porro II's as well I'm fairly sure) had cemented baffled three part prisms and the pre-war Zeiss Turita/Telita line of roof prisms had cemented baffled two part prisms. But the circa WW II Ross, Barr & Stroud, and Huet Porro II's (all of which were two part prism assemblies excluding  the field lens) for some reason were not baffled.

Leica Kern AARAU cemented Porro I prisms: https://www.flickr.c...v-aPP4zz-bua4hR

Bell & Howell M19 7x50 cemented Porro I prisms: https://www.flickr.c...F-9dWwuW-bzDGax

Leitz (beh) Kriegsmarine 7x50 cemented and baffled three part Porro II prisms: https://www.flickr.c...n-pzYVyt-9v9jiB

Zeiss Turita 8x24 cemented and baffled roof prism: https://www.flickr.c...r-8WbRsW-bu652J

Ross Binoprism No. 5 7x50 two part cemented prism without baffling between the prisms: https://www.flickr.c...7-csdVUw-fw4sxr


Edited by FrankL, 24 May 2016 - 09:30 AM.

  • Mad Matt likes this

#7 marktd

marktd

    Mariner 2

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

Posted 24 May 2016 - 09:57 AM

 

Leica Kern AARAU cemented Porro I prisms: https://www.flickr.c...v-aPP4zz-bua4hR

 

 

I think this illustrates that mounting cemented prisms can be as simple, maybe even more simple, than using hypotenuse faces of uncemented prisms. 



#8 Mad Matt

Mad Matt

    Vanguard

  • *****
  • Posts: 2379
  • Joined: 20 May 2003
  • Loc: Frankfurt, Germany

Posted 24 May 2016 - 10:14 AM

I would speculate that before multi coating, binoculars with cemented prisms where noticably brighter. Now that prism faces are coated, it's cheaper to mount them individually on a prism plate that incorporates baffles.
  • Jon Isaacs and StarDustBin like this

#9 marktd

marktd

    Mariner 2

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

Posted 24 May 2016 - 10:31 AM

I'm not entirely sure it's cheaper to mount individually.  I can see that the typical assembly process is very simple and cheap -- purely mechanical, just slap the parts together, put in a couple screws.  

 

But is cementing any more costly?   Inexpensive objectives are cemented.  Clearly there's an efficient factory assembly process for this, and it has to be just as easy to cement a pair of prisms together.    At final assembly, the cemented prisms would then just be a part to slap together and screw down -- could be even fewer parts and screws.  The Kern example is epoxied and that's different, but it'd be easy enough to design a holder for a cemented pair that would have a single retaining strap and pair of screws.



#10 Mad Matt

Mad Matt

    Vanguard

  • *****
  • Posts: 2379
  • Joined: 20 May 2003
  • Loc: Frankfurt, Germany

Posted 24 May 2016 - 10:53 AM

Just visualize the process of applying glue to one prism face, securing the prism in the jig, carefully positioning the other prism on the face ensuring that there are no air bubbles in the cement.  Then aligning with the jig and activating the UV curing light for the prescribed time (probably 10-20 seconds). Lastly, the prism must be glued to the mount, which itself is also a more complex structure because there reference surface is not the hypotenuse surfaces.

 

Now visualize the process of taking a single prism, placing it in the machined recess that the more simplistically designed prism plate has, putting two dabs of hot glue on it, flipping it over and repeating the process.

 

Which process do you think will provide more consistent results when performed repetitively and as quickly as possible?


Edited by Mad Matt, 24 May 2016 - 10:58 AM.

  • StarDustBin likes this

#11 marktd

marktd

    Mariner 2

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

Posted 24 May 2016 - 11:44 AM

Well, I've actually done it, and I can easily visualize it in a production assembly line -- I have a little background in manufacturing and have designed and assembly built jigs and such, etc.  Even with very low-tech methods it looks like it could easily be done in well under a minute per pair (excluding cure time).  And it's no more complicated than cementing lens -- just different jigs.  And the prism does not have to be glued to the mount. As I noted, one can easily design a mount that holds cemented prism pair with screws, straps, etc.  



#12 GlennLeDrew  Happy Birthday!

GlennLeDrew

    Hubble

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

Posted 24 May 2016 - 05:20 PM

The Porro I 'middle' baffle aperture is virtually always smaller than the entrant/exit baffles. This bigger-smaller-bigger baffle aperture arrangement is permissible because of the geometry defined by the tapering inward of the image-forming light cones and the tapering outward of the chief rays on their way to the field stop edge. The middle baffle's principal task is to reduce the passage of light scattered by the first prism's interior walls, which surfaces are more brightly illuminated by out-of-field light entering the objective.

 

If a 'painted' or sheet material middle baffle can be successfully cemented, then fair enough. And we see that such has been done in a *few* systems. But I've never seen an example myself, which implies the large scale omission must represent a significant enough labor/cost savings.

 

Unless a cemented prism pair is supported at both prisms, cantilevering results, which is less robust against shock and high-G acceleration.


  • StarDustBin likes this

#13 marktd

marktd

    Mariner 2

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

Posted 24 May 2016 - 09:56 PM

So this bigger-smaller-bigger is Porro I specific, and does not apply to Porro II?  As you noted, Porro II are normally cemented, but I don't see any examples with baffles at the cemented surfaces.

 

As you note, the scarcity of cemented Porro I has to have some basis.  There's likely some "good" reason, though it's also entirely possible it's a somewhat arbitrary convention that just got established and copied.  Not every manufacturing convention has an entirely rational engineering or economic justification underlying it. 

 

Also mysterious however, is that there's so few example in premium glass (pre-coatings era), where marginal added manufacturing cost would not be a factor, especially when weighed against the benefits of eliminating two air/glass surfaces. 

 

In any event, I was really wondering if there was some good optical reason not to do it.  I've cemented a couple and don't see any issues, but I've not performed any testing other than casually gazing through them.  But I am going to make something for real use now and thought I'd check the optical expertise here to see if I was missing something important.



#14 MartinPond

MartinPond

    Aurora

  • *****
  • Posts: 4948
  • Joined: 16 Sep 2014

Posted 24 May 2016 - 10:48 PM

I'm assuming you mean the Porro I type, for the Porro II very commonly has its three parts cemented.

 

Cementing the two glass blocks would complicate their supporting, by losing the handily large, flat, hypotenuse faces. Moreover, this would eliminate the fairly important baffle aperture at the interface between the two prisms.

Bingo.

The big faces on the large 'shelves' are very easy to make and tweak.



#15 marktd

marktd

    Mariner 2

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

Posted 24 May 2016 - 11:12 PM

Once I invested 5 minutes in a demo jig, it took ~30 seconds to cement prisms, start to finish.  

 

Lay first prism on V; put drop of cement on face; lay second down; press against alignment stops; hit with UV.  

 

Bingo. 

 

(note third axis alignment stop not shown; removed to take picture)

 

 

Attached Thumbnails

  • IMG_3978.jpg


#16 MartinPond

MartinPond

    Aurora

  • *****
  • Posts: 4948
  • Joined: 16 Sep 2014

Posted 24 May 2016 - 11:44 PM

Works by itself...

 

Supporting and aligning that inside a metal chassis would be a lot trickier

than two prisms on shelves, though. There is also shock: the prisms with

spring clips on shelves can usually take a small drop without losing alignment

or breaking. The cemented pair would generate a lot of shearing force.

With the right plastic 'carriage' something could be done to make it work, though.

 

I won't question the cementing: the dielectric difference is small,

and the loss across that tiny cemented gap is likely even lower than

the Canada Balsam in doublets.



#17 GlennLeDrew  Happy Birthday!

GlennLeDrew

    Hubble

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

Posted 25 May 2016 - 10:12 AM

Mark,

In order to rely on a fixed jig to position prisms for cementing it would require that the prisms be very accurately shaped. Real prisms typically have errors in one or both the angles of the faces (45-90-45) and the tilt that results in a very gently sloping section of a pyramid, which usually necessitates some small rotation to eliminate 'lean', or a departure from 180 degrees in image rotation.



#18 Gordon Rayner

Gordon Rayner

    Gemini

  • *****
  • Posts: 3255
  • Joined: 24 Mar 2007

Posted 25 May 2016 - 05:03 PM

Hanna in ATM II or III, discussed prism and prism seat tolerances in detail. Willmann-Bell, Richmond, Va. USA.

#19 MartinPond

MartinPond

    Aurora

  • *****
  • Posts: 4948
  • Joined: 16 Sep 2014

Posted 25 May 2016 - 06:29 PM

They typically compensate for the errors Glenn is talking about with a little

foil tape tab on the shelf (if anything)..   It's easy to 'rough them in' that way.

 

The gluing would work fine for a telescope. It's a much fussier

tolerance lining up binocular sides.



#20 marktd

marktd

    Mariner 2

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

Posted 25 May 2016 - 10:43 PM

Indeed. 

 

I've disassembled a bunch of binos recently and notice some have provisions for small alignment adjustment, and some don't.  Mostly related to quality and power I assume.  A pair of Bushnell Insta-Focus (clearly cheapie) have no means to adjust prisms, but a 60's Sears does (were those cheapie, or low/mid-range in their day?)

 

Interestingly, SARD Mark 21 has no means to adjust.  Judging by appearance alone, the exterior of the prisms appear much more precisely ground than the prisms of the old Japanese cheapies.   Maybe they were ground to tolerances sufficient to meet the needs (without adjustment) of a 7x war glass?

 

In any event, I now see that optically precise assembly requires an optical alignment jig (whether cemented or mechanically restrained), not just a mechanical jig.

 

Do you know of a standard or rule of thumb regarding required alignment precision?  



#21 MartinPond

MartinPond

    Aurora

  • *****
  • Posts: 4948
  • Joined: 16 Sep 2014

Posted 25 May 2016 - 11:19 PM

"

Do you know of a standard or rule of thumb regarding required alignment precision?

"

 

According to a character known as "WJC",

The US Navy had a repair standard of 2-4 arc-minutes coming out the eyepiece, but that's felt to be

    way tighter than the eye can see.   Russians found 30-40 arc-minutes coming from the eyepiece

   was a fair tolerance.

 

As you pull your eyes away from the eyepieces (on binoculars that don't 'black out'),

  the deviation increases (in the amount center field differs from your close-in IPD).

  A handy way to do a quick check (with an object on a plain background)..



#22 SMark

SMark

    Skylab

  • *****
  • Posts: 4046
  • Joined: 29 Aug 2011
  • Loc: Atlanta, GA USA

Posted 26 May 2016 - 12:46 AM

According to a character known as "WJC",

The US Navy had a repair standard of 2-4 arc-minutes coming out the eyepiece, but that's felt to be

    way tighter than the eye can see.   Russians found 30-40 arc-minutes coming from the eyepiece

   was a fair tolerance.

 

As you pull your eyes away from the eyepieces (on binoculars that don't 'black out'),

  the deviation increases (in the amount center field differs from your close-in IPD).

  A handy way to do a quick check (with an object on a plain background)..

 

I think you meant to say... According to a curmudgeon known as "WJC"...,

 

:coolnod:


  • Albie likes this

#23 GlennLeDrew  Happy Birthday!

GlennLeDrew

    Hubble

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

Posted 26 May 2016 - 02:21 AM

Martin,

Pulling your eyes farther back from the eyepieces has no impact on *angular* divergence, or image separation. An angle is an angle is an angle. The increasing *pysical* separation via *projection upon a surface* is distance dependent, but some angular divergence of, say, 0.5 degrees, is the same whether the eye is in contact with the eyepiece, at the eye point, 5cm farther rearward, or 50cm rearward.

 

The farther rearward eye placement only restricts the FOV, which might make the image displacement *appear* to be larger. But rest assured, the angular divergence is quite unchanged. If at the correct eye placement I have to cross my eyes a certain amount in order to fuse the images, that very same amount of eye crossing will apply at any other distance, as long as the FOV has not shrunk so small as to not permit the seeing of both images simultaneously.


  • Pinac likes this

#24 MartinPond

MartinPond

    Aurora

  • *****
  • Posts: 4948
  • Joined: 16 Sep 2014

Posted 26 May 2016 - 05:44 AM

You said:

"the angular divergence is quite unchanged."

 

Yes, absolutely!

 

And thus, the physical divergence of your view gets bigger and bigger.

 

 

It's time to take out your laser.

 

I first discovered the concept while using a laser to find the deviations.

 

If I face the objective directlyfrom 100 ft away (with the binocs pointed and focused on its origin site),

and wiggle the laser across the objective, a disc can be 'painted' on a borad behind the binoculars.

 

If the collimation is off, the L + R discs (as drawn by a vounteer with a mark following dots)

  show very distinctly-deviating locations.   The divergence increases as you move the board back.

 

Now:

  if I move my eyes back from the EPs,

  I see a smaller and smaller piece of the image, in good focus.,

  and if there is deviation, I end up looking at different pieces of the field.

  Smaller bits of a different image.

 

It's too late to block either of those phenomena with lore.

It really happens and I began with empirical work.  I even corrected a few

units with the laser-drawn discs, but it's too cumbersome now.



#25 marktd

marktd

    Mariner 2

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

Posted 26 May 2016 - 09:46 AM

I'm familiar with the collimation standards in general, as well WJC, who announces his curmudgeoness by starting his post with:

 

"the average observer on Birdforum is a few shades more mature and cranial than the average observer on Cloudy Nights.." 

 

What I actually was asking is if there's any known standard for prism rotation itself.  I understand this would not be a binocular end-state standard, but rather an standard internal to the manufacturing process for this particular sub-assembly.

 

In the samples I've disassembled, the prism rotation is only adjustable when the prism sub-assembly is outside the binocular housing, so you can't adjust rotation as part of the final collimation, during which one can adjust objective eccentrics.  This strongly suggests that during manufacturing, the prism sub-assembly is treated as a discreet unit with independent tolerances.  It would be very impractical and inefficient to check collimation, then partially disassemble the binocular, remove the prism sub-assembly, adjust rotation, re-install prism sub-assembly, re-test collimation, etc.

 

I understand this is a very specific and obscure question, and few, if any, outside the manufacturing process designers would likely know.  It's analogous to asking about machining tolerances on a binocular part.   But hey, there are some folks here with obscure expertise, can't hurt to ask!




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