View Full Version : Indexed Cams on RF Lenses and Adapters
This thread is intended for discussion on the need to introduce an indexed Cam to adapt lenses made for the Zeiss Contax for use on the Leica. It carries over the discussion started on this:
To perfectly match the an Adapter made for using Zeiss lenses on the Leica RF, a small cut needs to be made in the Cam to correct for the 0.1mm difference. Roland (Ferider) did this with an Adapter, and it works well.
The original Orion adapters did not make this cut. I do NOT make this correction on my J-3's, J-8's, and Zeiss Sonnars when I use them on the Leica and Nikon. I shim the lenses for optimal use close-up and wide-open. This produces an error at the infinity end of the focus range. I stop the lenses down for infinity use, and the Sonnar focus shift pushes the correct focus backwards. This and the DOF produces perfectly usable results.
BUT: to attain optical performance for wide-open work over the full focus range, a cut needs to be made into the cam OR the actual focal length of the lens must be reduced. I have done the latter a couple of times.
Roland's modified M-Mount adapter:
My Type 2 Rigid Summicron-M does the same thing, but the index cut is even more shallow.
Ny Nikon S2, shimmed for use with Zeiss lenses.
I moved the helical to acheive perfect focus with the Zeiss Sonnar on it for 8ft.
It's fine wide-open and close up, and infinity at F2.8.
My 13.5cm F3.5 for Contax mount can be used with it wide-open out to at least 30ft.
Contax mount lenses rotate 270degrees moving from 3ft to infinity. Nikon S-Mount lenses roate 260degrees moving the same distance. The pitch of the helicals is the same. This is responsible for the focus error between the two. I could have cut an indexed cam into the S2, but the Shim works just fine. The residual error is covered by F1.5 DOF close-up, and covered by the Sonnar focus shift for infinity.
So: How much is close-enough? I use a second Zeiss Sonnar on a "Nikon-Leica" Amedeo adapter. I custom shimmed the Sonnar 50/1.5 to the adapter. It works perfectly. An indexed cam would also work, and perhaps be more accurate. It would preclude the use of my Nikon external mount lenses on the adapter, or at least move the inaccuracy to external mount lenses.
To illustrate the differences in focus across range:
1953 KMZ J-3 in LTM , shimmed for the Leica, at F1.5
Focus is on the distant tree line, using the Canon P RF for focus. The best focus is slightly in front of it. Stopping down to F2.8 would shift focus backwards, or I can just move slightly past what the RF indicates in "actual use".
Same lens, wide-open and close-up. Focus is spot-on.
I shoot with it, and am very happy with the results.
Let me get this straight; if I do not plan to use any Nikon lenses with the Contax-Leica adapter, then having the extra .1mm shim. If I use my J-3 in Contax mount with the adapter without the extra .1mm shim, then the lens will not be focusing perfectly when used wide open. If the adapter gets a "cut", then the lenses would focus accurately both for close-focus and at infinity but the adapter would become useless with Nikon lenses.
Without a cut in the adapter, is then my J-3 optimized for "average use" at middle distanced when used wide open?
Does this mean that maybe one J-3 could be labeled for Leica use and for close focus, and only that lens be shimmed extra?
Yes, that is pretty much it.
HOWEVER: In real life, shimming the Zeiss standard lens for close-up and wide-open use works for me. Distance shots are often stopped down. BUT: using a Zeiss external mount lens on the adapter, for best work from closest focus to infinity, the 0.1mm cut in the RF cam is best. This cut is NOT required for using Nikon external mount lenses with the Leica.
Most Leica lenses will full cam appear to have been hand-indexed for final calibration. While barely visible, my Summicron 90 v3, Summilux 50 v2, and Summilux 75 have a slightly non planar cam, with faint grind marks.
One more interesting application I thought about, Brian: one could index the cam of a googled Summaron or Summicron to get rid of the googles.
Just never tried it ...
I am curious about why your Leica Summicron 50 has that little bit milled away from part of the cam. You said you had to do it to Contax lenses to get them to focus right on a Leica because of the difference in the focal length each system's RF was designed for, but a Leica lens should have been designed to work on a Leica body without the cam being milled.
I ask because when I bought my 50mm Summicron, which is a black tabbed version, I got it from KEH and the first one they sent me had part of the cam milled away like yours. It backfocused several inches up close! I returned it and they sent me another copy of the same version Summicon and that one did NOT have part of the cam milled off....and it focuses PERFECTLY at all distances.
The Leica RF is calibrated for 51.6mm. But it is very difficult to make each lens by exactly 51.6mm, and Leica is no exception. My Summicron is 51.9mm. The indexed cam is to correct that small difference. Leica could do this for the M-Mount as the lens locks in place, and the Cam will always come to the same beginning point. For Screw Mount lenses, indexed cams are more difficult to implement as the starting position must always be the same. I suspect one reason for the move to M-Mount was to allow more optical variation that could be corrected by indexing the Cam.
With your lens being off: either the Optics module was mismatched to the Focus Mount or it was shimmed incorrently.
As machinist I can suspect the situation where you need a helical screw pitch of 4.64mm for a specific lens, but the lathe or machine tool used only let you use 4mm, 4.5mm or 5mm, thus you need mill the second helical with a pitch of 0.14mm and use the 4.5mm option in the machine tool for to complete the 4.64mm screw pitch. Also is valid 5mm and a correction of 0.36mm ( first case clockwise and the second counterclockwise)
With the modern numerically controlled machine tools this not is necessary because the helical pitch is determined in the CNC software.
vBulletin® v3.6.8, Copyright ©2000-2013, Jelsoft Enterprises Ltd.