Switch to English Language Passer en langue française Omschakelen naar Nederlandse Taal Wechseln Sie zu deutschen Sprache Passa alla lingua italiana
Members: 70,571   Posts: 1,545,599   Online: 983
      
Page 1 of 2 12 LastLast
Results 1 to 10 of 11
  1. #1
    Marco B's Avatar
    Join Date
    Jan 2005
    Location
    The Netherlands
    Shooter
    Multi Format
    Posts
    2,983
    Images
    169

    My first lens repair...

    OK, so I had another accident, my camera bag toppled of the table at home and my Minolta Dynax 7 hit the ground with my relatively heavy and big Sigma 28-70 F2.8 EX attached to it. Opened the bag, to see it still attached to the camera body, however at a rather odd angle...

    Now some of you who are not particular fond of Sigma will say: well deserved, throw it out... but I actually have good experiences with this particular lens and I am not really in a financial position for big new sales or expensive repairs at the moment.

    So, since it is an older lens too, I decided to give a go and at least see what was broken inside. Fearing the worst, I started disassembling the lens from lens mount down, unscrewing stuff and carefully noting how it was assembled.

    It turned out to be easier than expected. Maybe this is also due to Sigma being a general manufacturer, creating lenses for different mounts and brands of cameras, but the lens mount unit is almost separate from the true lens unit. There is no direct and "fixed" connection between the lens mount and the lens unit. For example, the mechanical parts moving aperture and AF, that are transmitted from the camera body, are being moved in the lens by a lever and small axle that are not physically connected to the true lens assembly. Meaning any forces exerted on the lens mount in case of an accident like mine, will not damage the lens assembly itself. In fact, the AF axle falls right out of the lens when disassembled, it is only held in place by the screwed on lens mount.

    I removed a couple of parts, including the true metal lens mount, to end at what seemed to be the culprit, a slightly bend and broken plastic ring. No further damage was visible, and moving the aperture and focus levers and gear stuff, seemed to indicate the lens to be OK.

    The "ring" broken, was actually broken at a seemingly insignificant place, at a point where it narrowed from a wide strong ring to a minor 3 mm wide plastic "bridge". As a consequence of the forces exerted during the fall, the thick part of the ring had bend though, and now stuck out about 1 mm at one end. Trying to bend it back proved difficult, as at this part of the "ring", the ring was reinforced at the back with a kind of "tube" like extension, going about 7mm deeper into the camera.

    I managed to bend it back some, and reassembled the lens completely, to see if it would actually still work. Although it was way better, the lens still was at a minor odd angle to the camera body, but I could at least test it now. AF and aperture seemed fine!

    I than disassembled the whole unit a second time, trying to assess what could be done. I tried to remove the affected part by unscrewing three screws, but it seems this part is also glued in the lens assembly. So far for the option of removing it and sending it of to Sigma to ask for a replacement...

    I than decided to have a look if I could somehow bend the ring back and fix it. After some difficulties, I finally managed to bend it back some, and fix it in place with super-glue. It seemed right, but after reassembling the lens, it turned out to be still at a very minor odd angle to the mount. Just maybe a 0.2 to 0.3 mm gap at one side.

    I disassembled again, and noticed the other parts of the lens mount fitted onto the damaged ring using screws that fitted into tiny holes that were slightly raised compared to the rest of the "ring". Since it was just plastic, I took a piece of sand paper and started to take off some of the plastic of one of the three screw points at the point of breakage. When it seemed right, I reassembled the whole lens.

    Bingo! It now seems to be OK. Everything works fine, and the lens no longer stands at an odd angle...

    For the moment, it will have to do. Everything feels OK at least, no strange hiccups of parts indicating further issues. I think the breakage and bending of this plastic part took the main blow and forces off of the rest of the lens assembly.

    Anyway, one last question to you all. Of course, although it now visually seems OK, there may still be a very minor deviation from perpendicular. I started wondering how big an issue this would be? I have no idea how much "depth-of-field" there actually is at different apertures on the film plane in the camera. Will any minor deviation immediately show up? Or is there actually some "tolerance" in depth of field..., even for the biggest F2.8 aperture setting?

    I have been thinking of photographing a newspaper to see how the sharpness is across the film plane. Of course, this would require very careful setup, as I could just as well screw it up there.

    But anyone who can comment on this or has suggestions?

    Marco
    Last edited by Marco B; 01-31-2010 at 05:12 AM. Click to view previous post history.
    My website

    "The nineteenth century began by believing that what was reasonable was true, and it wound up by believing that what it saw a photograph of, was true." - William M. Ivins Jr.

    "I don't know, maybe we should disinvent color, and we could just shoot Black & White." - David Burnett in 1978

    "Analog is chemistry + physics, digital is physics + math, which ones did you like most?"

  2. #2
    ic-racer's Avatar
    Join Date
    Feb 2007
    Location
    Midwest USA
    Shooter
    Multi Format
    Posts
    6,225
    At 28mm, alignment to the micrometer-level would be required to have the image evenly in focus across the film plane.

    With the lens wide open, shoot a distant landscape and check the edge to edge horizontal image. My guess is that it may not be sharp all the way across. If it is acceptable, then you are set or you can flip the camera and check the top-to-bottom alignment the same way.

  3. #3

    Join Date
    Dec 2009
    Shooter
    4x5 Format
    Posts
    127
    Excellent work Marco!
    I discovered the advantage of plastic lens bodies years ago when I was a repairer. Once in a while the multistart threads (for focus) get crossed in an impact - with metal this would certainly need replacement. Often though with plastic I was able to simply force the two parts back into place and they would work perfectly. Just springs back to its original shape beautifully.
    Early plastic mouldings could be a problem - it seemed to me at the time that designers were still thinking in terms of machined metal - plastic offered new design opportunities but it took a while to figure out how best to use the new materials.
    Many people still regard plastic as second-rate for this purpose and many of the more expensive lenses still use alloy. Plastic though will often survive a fall which would write-off a metal lens.
    Sure, check the alignment with a newspaper but if it isn't 100% and works well otherwise - how often would it matter for everyday photography?
    Well done, take some pictures!

  4. #4
    Marco B's Avatar
    Join Date
    Jan 2005
    Location
    The Netherlands
    Shooter
    Multi Format
    Posts
    2,983
    Images
    169
    Quote Originally Posted by ic-racer View Post
    At 28mm, alignment to the micrometer-level would be required to have the image evenly in focus across the film plane.
    Thanks ic-racer, that's the kind of info I wanted to hear. However, the micrometer level has me slightly baffled though. Even the film's emulsion is already a dozen or so micrometers thick if I remember well. And looking at the lens mount assembly, with all kinds of stacked stuff, it even makes me wonder how they would achieve that level.

    Can you answer a few questions:

    Is the micrometer level only necessary for wide open (F2.8), or is it less of problem with more regular aperture settings like F5.6 / F8. Maybe a stupid question, but does depth of field on the filmplane increase with larger F stop like the regular depth of field in front of the lens does?

    Quote Originally Posted by ic-racer View Post
    With the lens wide open, shoot a distant landscape and check the edge to edge horizontal image. My guess is that it may not be sharp all the way across. If it is acceptable, then you are set or you can flip the camera and check the top-to-bottom alignment the same way.
    Can you elaborate a bit more:

    - What is the advantage of the distant landscape over the "newspaper" test?
    - Do you mean focus at infinity, F2.8 setting and 28mm?
    - Why flip the camera? Do you mean a vertical shot, and shouldn't I be able to judge the same thing from a single horizontal shot, or is it because you want the line of the horizon central across the image to check for focus along that line?

    Quote Originally Posted by unclemack View Post
    Sure, check the alignment with a newspaper but if it isn't 100% and works well otherwise - how often would it matter for everyday photography?
    Well done, take some pictures!
    Thanks, I will do both, to be sure I am at least in the ballpark of proper alignment.
    My website

    "The nineteenth century began by believing that what was reasonable was true, and it wound up by believing that what it saw a photograph of, was true." - William M. Ivins Jr.

    "I don't know, maybe we should disinvent color, and we could just shoot Black & White." - David Burnett in 1978

    "Analog is chemistry + physics, digital is physics + math, which ones did you like most?"

  5. #5
    Marco B's Avatar
    Join Date
    Jan 2005
    Location
    The Netherlands
    Shooter
    Multi Format
    Posts
    2,983
    Images
    169
    Hi all,

    Well, I guess there may be a little bit more work ahead. I came up with another idea to "test" the lens (still plan on the newspaper shooting, but that requires processing and more work first). I put the lens on a solid flat surface (base board of my heavy Durst L1200 enlarger which is rock solid), got my bubble level, took a "base reading" of the level of the base board, and next balanced it on the top of the 72 mm thread on the top of the lens standing straight up. While keeping the bubble level in the same position, I rotated the lens under it and looked if and how the bubble changed position. It did, which in perfect conditions it wouldn't.

    Next, I did a rough "g(uess)estimate" of the error in levelling based on the displacement of the bubble I saw. I came up with something in the range of about 1-2.5mm / meter length, indicating an error of about 0.03-0.09 mm (30 to 90 micrometers) across a 36 mm wide film plane.

    That seems a bit much... although my next question is of course: at what F stop do you have at least a 100 micrometers of depth of field at the film plane (probably also depended on focusing distance)?

    Looking through the viewfinder, the image does seem to be reasonably in focus across the frame, but this is probably not a particular good indicator of the true situation... just like the visual appearance of the lens that seems about OK but clearly is not yet.

    Anyway, I will try to improve this by disassembling the lens again and see what I can do before trying to shoot with it.

    Marco
    My website

    "The nineteenth century began by believing that what was reasonable was true, and it wound up by believing that what it saw a photograph of, was true." - William M. Ivins Jr.

    "I don't know, maybe we should disinvent color, and we could just shoot Black & White." - David Burnett in 1978

    "Analog is chemistry + physics, digital is physics + math, which ones did you like most?"

  6. #6
    bobwysiwyg's Avatar
    Join Date
    Mar 2008
    Location
    Ann Arbor, MI U.S.A.
    Shooter
    Multi Format
    Posts
    1,555
    Images
    3
    Good luck with the repair. I have the very same lens and love it actually. My only complaint is the heft.. which I got over some time ago.
    WYSIWYG - At least that's my goal.

    Portfolio-http://apug.org/forums/portfolios.php?u=25518

  7. #7
    ic-racer's Avatar
    Join Date
    Feb 2007
    Location
    Midwest USA
    Shooter
    Multi Format
    Posts
    6,225
    Some numbers from my experience aligning the front standard detents on my 6x9cm view camera with 65mm lens (gives the exact same field of view as a 28mm lens on a 35mm camera, but just about twice as big.)

    I could detect on film a deviation of the lensboard of two thicknesses of electrician's tape. So 10 layers of that are about 2.5mm, so 0.25mm each piece. So half of that is about 0.1mm or 100 micrometers.
    Last edited by ic-racer; 02-01-2010 at 08:37 PM. Click to view previous post history.

  8. #8
    ic-racer's Avatar
    Join Date
    Feb 2007
    Location
    Midwest USA
    Shooter
    Multi Format
    Posts
    6,225
    Quote Originally Posted by Marco B View Post
    That seems a bit much... although my next question is of course: at what F stop do you have at least a 100 micrometers of depth of field at the film plane (probably also depended on focusing distance)?
    The view camera focusing equation works here. Solve for 0.1mm as the focal spread.
    http://www.largeformatphotography.info/fstop.html

  9. #9
    Marco B's Avatar
    Join Date
    Jan 2005
    Location
    The Netherlands
    Shooter
    Multi Format
    Posts
    2,983
    Images
    169
    Quote Originally Posted by ic-racer View Post
    Some numbers from my experience aligning the front standard detents on my 6x9cm view camera with 65mm lens (gives the exact same field of view as a 28mm lens on a 35mm camera, but just about twice as big.)

    I could detect on film a deviation of the lensboard of two thicknesses of electrician's tape. So 10 layers of that are about 2.5mm, so 0.25mm each piece. So half of that is about 0.1mm or 100 micrometers.
    Thanks ic-racer.

    So you're actually saying that from practice you were able to see any deviation >= 0.5 mm degrading the image on film (6x9 film format). Which doesn't sound strange to me.

    So the 30-90 micrometer deviation is clearly at the limits of acceptable... still need to do more work to get it better than (which isn't a surprise), or turn it in for a true professional repair if I don't succeed in getting it better...

    In addition, I now took out my Sigma 70-200 F2.8 EX lens, and did the same "bubble level" test. This lens is pristine and hasn't seen any such unfortunate accidents. As expected, the bubble level indicated near perfect perpendicular alignment of the lens, with hardly detectable movement of the bubble.

    Marco
    My website

    "The nineteenth century began by believing that what was reasonable was true, and it wound up by believing that what it saw a photograph of, was true." - William M. Ivins Jr.

    "I don't know, maybe we should disinvent color, and we could just shoot Black & White." - David Burnett in 1978

    "Analog is chemistry + physics, digital is physics + math, which ones did you like most?"

  10. #10
    Marco B's Avatar
    Join Date
    Jan 2005
    Location
    The Netherlands
    Shooter
    Multi Format
    Posts
    2,983
    Images
    169

    Some good news...

    Ok, some good news now. I managed to improve the alignment somewhat. Perfect is not possible without a professional repair.

    But...

    I now took the pragmatic approach unclemack already suggested, and starting shooting. Besides also making one control shot of a test chart with fine text at F2.8 and 28 mm wide angle, I also took it out for real shooting.

    After development of the roll, I first looked at the test chart with a 10x loop. Corner to corner sharpness seemed good. I than went on to print some of the negatives that had very fine detail in them, and frankly, I can't tell what the lens went through ... good overall sharpness. Actually, I had more trouble keeping the negatives flat in my glassless negative carrier, causing some blurring on a quick test print, but the final prints showed no apparent issues.

    So although it's not perfect, it seems good enough for the time being! Phew..., I am glad, saves me a few bucks for a new lens...

    Marco
    My website

    "The nineteenth century began by believing that what was reasonable was true, and it wound up by believing that what it saw a photograph of, was true." - William M. Ivins Jr.

    "I don't know, maybe we should disinvent color, and we could just shoot Black & White." - David Burnett in 1978

    "Analog is chemistry + physics, digital is physics + math, which ones did you like most?"

Page 1 of 2 12 LastLast


 

APUG PARTNERS EQUALLY FUNDING OUR COMMUNITY:



Contact Us  |  Support Us!  |  Advertise  |  Site Terms  |  Archive  —   Search  |  Mobile Device Access  |  RSS  |  Facebook  |  Linkedin