Canon 50mm f1.2 LTM Repair Option

[qqphotos]

Ok - yes, there is a shim, and quite a hefty one - 0.82 mm! between the optical block and the focusing mount. Not sure how I managed to miss that earlier, I think I was looking for something much smaller. If I can borrow an M10 for a few days I'll see if I can fabricate a few thinner shim substitutes to see if I can improve things. It is certainly an easy lens to disassemble, at least. It also appears that it's very sensitive to how it's assembled, so it may have been put together sloppily by whomever last messed with it.

 

[Sonnar Brian]

Which brings up another question - the copy I have produces sharp focus (well, as sharp as this lens ever gets at f/1.2) consistently about 10-12 cm closer than the rangefinder thinks is correct, when wide open and at 1.5 meters distance. I expected some discrepancy but this is wildly off by enough that I suspect either it's shimmed wrong or has been disassembled and put back wrong.

It looks like that Canon bodies' flange distance is the same 28.8mm as Leica's, and my Barnack bodies focus other lenses well enough, so that leaves me suspecting the lens needs to be de-shimmed (i think?) which is unfortunate since there aren't any shims in it.

0.136330222mm reduction in shim required to bring the lens in agreement assuming a 10cm error at 1.5m
I use the equation for focal length, plug in the actual focus and the distance as measured by the RF, compute the difference.

1/f=1/d+1/b
1/b=(1/f)-(1/d)
b=1/((1/f)-(1/d))
where f is 51.6mm, and d is the distance. Solve for b, back focus. I have this in Excel.

 

[dexdog]

0.136330222mm reduction in shim required to bring the lens in agreement assuming a 10cm error at 1.5m
I use the equation for focal length, plug in the actual focus and the distance as measured by the RF, compute the difference.

1/f=1/d+1/b
1/b=(1/f)-(1/d)
b=1/((1/f)-(1/d))
where f is 51.6mm, and d is the distance. Solve for b, back focus. I have this in Excel.

Good thing that my 50/1.2 is in good shape- I don't think I could polish a shim down to the eighth decimal place in mm! 😛 I might be able to do it in meters on a good day; or maybe not. because that is still pretty small.

 

[peterm1]

0.136330222mm reduction in shim required to bring the lens in agreement assuming a 10cm error at 1.5m
I use the equation for focal length, plug in the actual focus and the distance as measured by the RF, compute the difference.

1/f=1/d+1/b
1/b=(1/f)-(1/d)
b=1/((1/f)-(1/d))
where f is 51.6mm, and d is the distance. Solve for b, back focus. I have this in Excel.

Good to know how to calculate this though for me its a bit academic - without a proper, accurate Engineer's micrometer I am reduced to trial and error or rather lap and test. 🙂 But as it happens it is still useful and heartening as it confirms that less lapping is needed than I thought might be. As it also happens, I also have another project on which I can try this once again (having done it a few times before - with success but purely by trial and error.

I recently bought a fairly mid-range lens adapter for Nikon F to Leica M mount. (This will be used only with a mirrorless camera, not a rangefinder one of course as there is no rangefinder coupling). When I used it, it was clear that I could not reach infinity with any lens I mounted on it.

These adapters are simply built with essentially three parts - the base plate which has the M mount, the mid-section / main body/ spacer and the top plate which has the Nikon lens mount. It's a matter of removing the top / M mount plate (undoing 5 screws) and performing the lapping / testing procedure on the top of the main body till it's back within spec. This sort if issue very occasionally happens with adapters unless you by a top brand one - though in my experience, mostly the discrepancy is small enough to be ignored for practical purposes. No great risk is involved as it's not an expensive adapter (and I have already replaced it anyway) but it's still nice to be able to use it as backup if needed.

When I measured mine (using a simple vernier caliper) it looked as if the deviation from what is required is slightly over 0.1 mm. In other words similar to the 0.136330222mm reduction you mentioned in your example.

For anyone interested in performing this type of operation themself if they are stuck with a "bodgy" adapter, I calculated the required alteration to the adapter by using tables of correct lens flange distances (readily available online) to calculate how thick the adapter should be by subtracting the correct M mount flange distance from the correct F mount flange distance. This gave me the "target" to work to when performing my "surgery". I then measured the actual thickness of the adapter from flange to flange and discovered, as noted above it was a little over 0.1mm out of tolerance - which is the best accuracy I can reach with my relatively simple measuring tools.

BTW lapping around 0.1mm by hand is still a substantial amount of work to perform especially considering the accuracy
required and constant checking therefore needed as you perform the work.

 

[qqphotos]

This has been a bit of a project. I figured I should first disassemble and reassemble properly in case something was crooked or loose, and in fact this did change the plane of focus significantly, though it was still off by a bit. I needed additional spacing so I spent a lot of time cutting thin brass shims and experimenting.

With this lens wide open, it's fairly hard to define "in focus" absolutely critically so it's a bit of a subjective thing. At this point i've added 0.03mm to the existing 0.82mm and it's actually quite well in agreement with the rangefinder on my M2 (compared to "quite bad" before).

It front focuses a tiny bit at f/1.2 with focus shift pushing it to just about dead-on at f/2-2.8 and beyond that there's enough depth of field that it doesn't matter much. And of course there's field curvature and lots of other aberrations to contend with. There are so many sources of potential error at this scale that it seems like an exercise in frustration to aim for "perfect."



My local, poorly-lit corner store. Focused on the small bottles on the counter in front of the cash register.