Transplanting TC14E-III Optics Into A TC14-II Body . .

[PedroS]

Don't know how you have made such measurements, but mine have been made with a Mitutoyo with constant pressure and accurate until 0,005mm.
Mine measures across the mount in 8 different points always 24,11mm...

[jhinkey]

If these firgures are accurate, they imply the mount(s) are warped and well outside factory tolerance.

Yeah, I'll check it later as I said with a more accurate 0-1" micrometer.  However I've found this kind of non-parallelism common with many adapters, etc. with no correlation between degree of non-parallelism and cost.

As a side note, I'm working on trying to develop a method to quickly measure flange to sensor parallelism as a way to do a self check w/o having to send the body in to a shop.

[Erik Lund]

OK but we end up at about 0.6mm difference between the two versions, that's 6 pieces of normal printing paper, a lot,,,

However - I don't see the IQ that would make this worth doing,,, several hours dismantling the two then machining, positioning and centering and reassembly of one converter and one into the parts bin,,,

[Mongo]

Quickly taking a screen shot of Erik’s images (which hopefully are to scale with each other), Mongo overlaid one on the other.

1.4 III - dark rose coloured elements with grey body

1.4 II - pink coloured elements with dark body

You will see from the edited diagram that some things line up and some do not. The rear mount lines up on one side but not the other. This must be an error in the diagram because the rear mount is perfectly parallel to the body. So, there is some diagram error or distortion in the proportions or the scale somehow. It is difficult to know what parts may be just housing and not the mount itself.

In Mongo’s view, you must be able to:-

 1.   transplant the whole of the version III elements and groups exactly as they are.
2.    ensure that the front and rear elements of the converter are exactly the same distance from the front and rear mounts respectively as they had on the Version III converter

It looks like you can fit all the version III elements/groupings into the version II housing IF you can overcome the slightly smaller housing at the front of the converter. If you can, then, it seems possibly to get that front element the correct distance from the front mount. However, it looks like the rear element will protrude a little too much from its required distance to the rear mount.

A lot of this is pure speculation without having precise scale diagrams and measurements.

Last point .........is it really worth this incredible trouble

[jhinkey]

Quickly taking a screen shot of Erik’s images (which hopefully are to scale with each other), Mongo overlaid one on the other.

1.4 III - dark rose coloured elements with grey body

1.4 II - pink coloured elements with dark body

You will see from the edited diagram that some things line up and some do not. The rear mount lines up on one side but not the other. This must be an error in the diagram because the rear mount is perfectly parallel to the body. So, there is some diagram error or distortion in the proportions or the scale somehow. It is difficult to know what parts may be just housing and not the mount itself.

In Mongo’s view, you must be able to:-

 1.   transplant the whole of the version III elements and groups exactly as they are.
2.    ensure that the front and rear elements of the converter are exactly the same distance from the front and rear mounts respectively as they had on the Version III converter

It looks like you can fit all the version III elements/groupings into the version II housing IF you can overcome the slightly smaller housing at the front of the converter. If you can, then, it seems possibly to get that front element the correct distance from the front mount. However, it looks like the rear element will protrude a little too much from its required distance to the rear mount.

A lot of this is pure speculation without having precise scale diagrams and measurements.

Last point .........is it really worth this incredible trouble

You can't really use those cross sections for anything accurate.  Measurements will tell.
Only way to know is to modify a TC14E-III to accept lenses that have aperture index tabs and do an A/B comparison on your favorite lenses.

[Erik Lund]

Indeed, these are not to scale,,,

The aperture follower 'shadows' the lower part,,, on the Mark I and II

It's an E converter - No linkage at all so I don't think you can fool it,,,

[Øivind Tøien]

As a side note, I'm working on trying to develop a method to quickly measure flange to sensor parallelism as a way to do a self check w/o having to send the body in to a shop.

Please enlighten us if you find a way. I have in the past returned two bodies due to incorrect register distance/misaligned mount; it would be great to have a better way than repeated comparison of infinity points or focus points with focus ring taped in fixed position.
 

[jhinkey]

Please enlighten us if you find a way. I have in the past returned two bodies due to incorrect register distance/misaligned mount; it would be great to have a better way than repeated comparison of infinity points or focus points with focus ring taped in fixed position.

I'm not working on measuring the distance, but rather the variation in the distance - i.e., non-parallel sensor and mount surface.
Two ways are interferometric and laser proximity probe (which is related to the first).  The goal is to make these simple and cheap, though that might be tough . . .

[Øivind Tøien]

To get the absolute register distance would be complicated by the UVIR filter/dust shaker that could reflect light back, and also the thickness of those that would have to be subtracted if light is reflected off the actual surface of the sensor. Anyone know how Nikon service do these tests?

My second problematic case was just incorrect register distance in a refurbished D7100 body, which did not show any signs of misalignment as far as I could tell, but the latter can be very hard to judge based on resulting captures at infinity. My feeling is that very short focal length lenses are more sensitive to misalignment of the mount than longer lenses due to the small movements necessary to effect focusing.

[jhinkey]

To get the absolute register distance would be complicated by the UVIR filter/dust shaker that could reflect light back, and also the thickness of those that would have to be subtracted if light is reflected off the actual surface of the sensor. Anyone know how Nikon service do these tests?

My second problematic case was just incorrect register distance in a refurbished D7100 body, which did not show any signs of misalignment as far as I could tell, but the latter can be very hard to judge based on resulting captures at infinity. My feeling is that very short focal length lenses are more sensitive to misalignment of the mount than longer lenses due to the small movements necessary to effect focusing.

Yes, measuring the absolute distance is very very difficult to do accurately - no idea how Nikon does it - likely indirectly.  Thus I'm just looking for non-parallelism . . . . much easier (but maybe not easy) to do.