120 Posted May 12, 2013 Share #1 Posted May 12, 2013 Advertisement (gone after registration) Has anyone tried this problem: Get an expression for the cam profile on the back of the lens. (Most of the accommodation for various focal lengths will come from the rest of the mount, but any lens, even a 50, will need a cam for exact focusing.) Apparently the Leica is spec'd by a 3 mm travel of the roller from infinity to 1m. Thus, close to the extension of a 50mm lens at 1m. Is there some official source for this? Where does the sometimes quoted 51.6mm spec come from--the 50 Elmar? Why 51.6 and not 51.8? Supposedly the lens set in the forties had cams no greater than 0.007 inches. Is this rather the cam required for a circa 50mm lens at 1m??? Thanks for any info or sources... Link to post Share on other sites More sharing options...
Advertisement Posted May 12, 2013 Posted May 12, 2013 Hi 120, Take a look here cams, rangefinders, specifications. I'm sure you'll find what you were looking for!
andybarton Posted May 12, 2013 Share #2 Posted May 12, 2013 Don't know anything about cams, but the 50mm lenses often have their true focal length engraved by way of two digits engraved next to the "feet" engraving. Mine is 51.9mm Link to post Share on other sites More sharing options...
120 Posted May 12, 2013 Author Share #3 Posted May 12, 2013 Thanks. This "sorting and grouping" is mentioned in Osterloh. It would be nice to know if the cam grinding is similarly customized, or maybe it's just one-size-fits-all for a nominal focal length. But really I just wanted to know the basics and am having trouble finding this info anywhere. Link to post Share on other sites More sharing options...
Michael Geschlecht Posted May 12, 2013 Share #4 Posted May 12, 2013 Hello Everybody, The roller travel is the same from Infinity to closest focus regardless of the focal length of the lens. The focussing helical to roller movement is designed to operate at a ratio of 1 : 1 with a 50mm lens. The corrections on the focussing mount (ie: 19) are incremental adjustments made to mounts to accomodate for differences in actual focal lengths of lenses involved. If the engraved distnce on the lens is "3 meters" then the roller moves the same distance to the same point regardless of the focal length involved. It is the lens mount extension that is differential to the roller movement with focal lengths that are less than or more than 50mm. When a lens is longer (ie: 90mm) the elements are extended outward further & at a greater rate as you focus closer. When the lens is shorter (ie:35mm) the elements ar extended outward from the focal plane less at a slower rate as you focus closer. Best Regards, Michael Link to post Share on other sites More sharing options...
120 Posted May 12, 2013 Author Share #5 Posted May 12, 2013 To be clear, I am also asking what the "standard" is ... is it the extension of a 50mm lens, extension of a 51.6mm lens, 3.0mm of roller travel, or something else? Link to post Share on other sites More sharing options...
120 Posted May 12, 2013 Author Share #6 Posted May 12, 2013 ... The focussing helical to roller movement is designed to operate at a ratio of 1 : 1 with a 50mm lens. The corrections on the focussing mount (ie: 19) are incremental adjustments made to mounts to accomodate for differences in actual focal lengths of lenses involved. ... Michael Are you sure about that... I don't think it will be 1:1 because the turning of the rangefinder prism and the extension of the lens have both to be related to the travel of the roller. So instead of the roller travel being linear in the lens extension, you get a little leftover term you need to cancel with a cam. Even with a 50mm lens. The cam on a simple 50mm lens would be subtle and you wouldn't see it's there. It's my understanding the longer and shorter focal lengths have a differential thread and, again, a cam ground on the male part, and I would like to get more info on the focusing mount...but I don't think you need the details of the focusing mount to figure out the cam. Can you explain more about the 19 above? Where can I read more about the Leica focusing mounts? Thanks Link to post Share on other sites More sharing options...
luigi bertolotti Posted May 12, 2013 Share #7 Posted May 12, 2013 (edited) Advertisement (gone after registration) Hello Everybody, The roller travel is the same from Infinity to closest focus regardless of the focal length of the lens. The focussing helical to roller movement is designed to operate at a ratio of 1 : 1 with a 50mm lens. The corrections on the focussing mount (ie: 19) are incremental adjustments made to mounts to accomodate for differences in actual focal lengths of lenses involved. If the engraved distnce on the lens is "3 meters" then the roller moves the same distance to the same point regardless of the focal length involved. It is the lens mount extension that is differential to the roller movement with focal lengths that are less than or more than 50mm. When a lens is longer (ie: 90mm) the elements are extended outward further & at a greater rate as you focus closer. When the lens is shorter (ie:35mm) the elements ar extended outward from the focal plane less at a slower rate as you focus closer. Best Regards, Michael Michael's excellent explanation gives, with the two phrases I evidenced, the basis for a possible computation of the rangefinder cam geometry : take the linear length of the cam of a 50mm lens that focuses 1m to infinity, it must have a gradient of 3mm over its total length : it is NOT a value that is the same for all 50mm lenses... it is well known that in modern lenses (say, a Summarit 50 of today), the linear length above quoted is shorter than in older lenses (say, a classic Summicron 50) , i.e., the so called "focus throw" is shorter... the gradient is higher. To answer simply to the OP, a "general expression" for the cam does not exist... or, better to say in math terms, it includes at least one parameter that is specific of the lens (I say "at least" because I do not know how is made the mechanical coupling of the cam to the focus helicoid... which in turn I do not know if has a fixed step in terms of millimeters x grade of rotation... in theory, with a certain coupling and a certain step, one could have a "flat" cam, with gradient=0) Edited May 12, 2013 by luigi bertolotti 1 Link to post Share on other sites More sharing options...
120 Posted May 12, 2013 Author Share #8 Posted May 12, 2013 ...To answer simply to the OP, a "general expression" for the cam does not exist... or, better to say in math terms, it includes at least one parameter that is specific of the lens (I say "at least" because I do not know how is made the mechanical coupling of the cam to the focus helicoid... which in turn I do not know if has a fixed step in terms of millimeters x grade of rotation... in theory, with a certain coupling and a certain step, one could have a "flat" cam, with gradient=0) I am assuming the step is fixed and lens is unit focusing. I think the only parameter needed that is "specific to the lens" is focal length. But I think you have to know exactly what the "standard" is to try to do the problem. If the answer is d.n.e., then that is something else. Link to post Share on other sites More sharing options...
luigi bertolotti Posted May 12, 2013 Share #9 Posted May 12, 2013 (edited) I am assuming the step is fixed and lens is unit focusing. I think the only parameter needed that is "specific to the lens" is focal length. But I think you have to know exactly what the "standard" is to try to do the problem. If the answer is d.n.e., then that is something else. Yes... indeed the "d.n.e." is a bad way to express my considerations... as I wrote before, I'd say is a parametric expression... (cams ARE designed, so, at the end, an expression to compute them MUST exist... they aren't designed by "trial and error"... ) ; thinking better at the example about the 50mm of different focus throw, a formula that gives, for a certain focus throw (= cam linear length) the gradient of the cam for a 50mm can be uniquely defined; I can't say quickly (but ought to be not to difficult to compute, having the time to think of... ) how the parameter "focal length" enters in the expression : I suppose it is in the form "Focal length/50" or vice-versa... as a multiplying factor of cam's gradient... don't know if with linear or more complex relation... studying and measuring directly the distance scales of lenses of different focals can be the starting point.... Edited May 12, 2013 by luigi bertolotti 1 Link to post Share on other sites More sharing options...
Michael Geschlecht Posted May 12, 2013 Share #10 Posted May 12, 2013 (edited) Hello 120, Welcome to the Forum. The "19" that I wrote was indicating 19 tenths of a millimeter. As explained in Andy's Post #2 above. These 2 numbers are the last digits of the actual (not nominal) focal length of that specific lens measured to the closest tenth of a millimeter. Without a decimal point between the whole millimeter & the tenth. Many, but not all of, Leitz/Leica lenses have these last 2 digits of their actual focal length engraved on their focussing mount without the decimal point. Some lenses do not. Divisible lenses often have that 2 digit number along with part of their serial number engraved inside on the separable lens head. The separable focussing mount often has the serial number of the lenshead which has been fitted. There are many exceptions & anomolies with Leitz/Leica. Best Regards, Michael Edited May 12, 2013 by Michael Geschlecht Link to post Share on other sites More sharing options...
Michael Geschlecht Posted May 12, 2013 Share #11 Posted May 12, 2013 Hello Again 120, As per focus travel with various focal lengths: Please re-read the last paragraph in my Post # 4 above. The original Leitz built in rangefinder of 1932 established the working relationship between roller travel & lens extension. That was: For every 1 millimeter of lens head travel toward or away from the film plane with a (nominal) 50mm lens there was 1 millimeter of equivalent roller movement toward or away from the film plane. When Leitz produced the M3 (the first "M" camera) it maintained this relationship. This allowed earlier lenses to be focussed with the M3's combined range/viewfinder. The earlier camera had a separate viewfinder. Best Regards, Michael Link to post Share on other sites More sharing options...
120 Posted May 13, 2013 Author Share #12 Posted May 13, 2013 Hello 120, Welcome to the Forum. The "19" that I wrote was indicating 19 tenths of a millimeter. s explained in Andy's Post #2 above... Thanks, didn't realize you were referring to post 2... Link to post Share on other sites More sharing options...
120 Posted May 13, 2013 Author Share #13 Posted May 13, 2013 ...The original Leitz built in rangefinder of 1932 established the working relationship between roller travel & lens extension. That was: For every 1 millimeter of lens head travel toward or away from the film plane with a (nominal) 50mm lens there was 1 millimeter of equivalent roller movement toward or away from the film plane. When Leitz produced the M3 (the first "M" camera) it maintained this relationship... Michael Thank you for that; that makes sense. Did you find it in a book, service manual, patent? This is 1:1, but in what I was describing, you can't have this and have the turning angle of the prism work out. You have to compensate somewhere and I thought it was done with a cam. I may be describing a simpler rangefinder and there is something more elaborate in the Leica threading or in the mechanism that turns the prism. The only number for cams I got from a Leitz visitor's report immediately after the war: "To ensure lineality between movement of the lens and the coupled telemeter a slight cam is machined on the end of the male member of the multi-start thread which engages with the roller which actuates the swinging prism of the telemeter. The greatest depth of this cam face does not appear to exceed 0.007 in." [regarding 50mm lenses]. So I tried and got 0.006 - 0.007" for the cam needed for 50 - 56mm lenses at 1m focusing distance, but for sure may just be coincidental. Link to post Share on other sites More sharing options...
giordano Posted May 13, 2013 Share #14 Posted May 13, 2013 This is 1:1, but in what I was describing, you can't have this and have the turning angle of the prism work out. You have to compensate somewhere and I thought it was done with a cam. I may be describing a simpler rangefinder and there is something more elaborate in the Leica threading or in the mechanism that turns the prism. I have always understood that the travel of the rangefinder roller between ∞ and 1m exactly matched the extension (focusing travel) of the ideal "5cm" lens (which as far as I can make out actually had a focal length of 51.6 or 51.8mm), precisely so that the commonest lens (the 5cm/3.5 Elmar) didn't need any sort of shaped cam, just the squared off end of the male focusing thread. Strictly speaking, "travel of the roller" means "displacement of a plane parallel to the film plane and tangent to the front of the roller". The angular movement of the swinging arm that carries the roller has a non-linear relationship with the displacement of the plane, but this is compensated for in the linkage between the swinging arm and the moving prism. The only number for cams I got from a Leitz visitor's report immediately after the war: "To ensure lineality between movement of the lens and the coupled telemeter a slight cam is machined on the end of the male member of the multi-start thread which engages with the roller which actuates the swinging prism of the telemeter. The greatest depth of this cam face does not appear to exceed 0.007 in." [regarding 50mm lenses]. The basic lens equation shows that a 51.8mm lens lens focuses to 1m with approximately a 2.83mm extension, and that a 0.1mm difference in focal length means approximately 0.01mm difference in extension. To put this in context, the 5cm/3.5 Elmar has depth of focus of about 0.23mm at infinity. A few hundredths of a millimetre aren't going to make any difference there. It's only with fast lenses that you need a profiled cam on a "50mm". At f/1.5 the depth of focus is less than 0.1mm so hundredths matter. Also, the more complex the lens design, the more its focusing travel will diverge (again, we're talking hundredths of a millimeter) from that predicted by the simple lens equation). 1 Link to post Share on other sites More sharing options...
jankap Posted May 13, 2013 Share #15 Posted May 13, 2013 There are cameras, which allow the user to optimize the rangefinder mechanism by DIY. See: Making a rangefinder cam for the Pacemaker Speed or Crown Graphic Jan Link to post Share on other sites More sharing options...
120 Posted May 13, 2013 Author Share #16 Posted May 13, 2013 I have always understood that the travel of the rangefinder roller between ∞ and 1m exactly matched the extension (focusing travel) of the ideal "5cm" lens (which as far as I can make out actually had a focal length of 51.6 or 51.8mm), precisely so that the commonest lens (the 5cm/3.5 Elmar) didn't need any sort of shaped cam, just the squared off end of the male focusing thread. Strictly speaking, "travel of the roller" means "displacement of a plane parallel to the film plane and tangent to the front of the roller". The angular movement of the swinging arm that carries the roller has a non-linear relationship with the displacement of the plane, but this is compensated for in the linkage between the swinging arm and the moving prism. The basic lens equation shows that a 51.8mm lens lens focuses to 1m with approximately a 2.83mm extension, and that a 0.1mm difference in focal length means approximately 0.01mm difference in extension. To put this in context, the 5cm/3.5 Elmar has depth of focus of about 0.23mm at infinity. A few hundredths of a millimetre aren't going to make any difference there. It's only with fast lenses that you need a profiled cam on a "50mm". At f/1.5 the depth of focus is less than 0.1mm so hundredths matter. Also, the more complex the lens design, the more its focusing travel will diverge (again, we're talking hundredths of a millimeter) from that predicted by the simple lens equation). O.k. so this is a different standard than Michael's, which was based on a 50.0mm lens. Yes, my "roller travel" was normal to the film plane. For your numbers you should take the distance (e.g. 1m) from the film plane, because here it matters (talking about cams). Using your standard (the extension of a 51.6mm lens) I get 0.0067" for the required 51.6mm lens cam. The fifty mm lens cam mentioned in the Leitz report was 0.007". To turn your numbers around to the other side of the lens, if a 0.007" cam is not there, and nothing else is changed, the focusing will be off about two inches at the 1m distance. In a nutshell you can't have the lens extension and the turning of the prism both linear in the roller arm travel, so you would have to compensate with a cam on the lens or somewhere between the roller and the prism. I guessed lens cam because that seems simpler, and also I got some vague hints in that direction, but I could be wrong. In any case I think it is a good problem, the same for anyone who did the Leica knock-offs after the war. Link to post Share on other sites More sharing options...
luigi bertolotti Posted May 13, 2013 Share #17 Posted May 13, 2013 Uhm... tonight I tried a bit of math (trigonometry, mainly...) based on the excellent schematics of Leica M rangefinder, published by Puts in his compendium..... (btw, the M rangefinder does allow to focus at 0,7 meters, which means that the 3mm movement of the roller of the Leica screw mount had been increased a bit) : I got a decent relation between distance of the subject to focus and linear movement of the roller (rather complex, of course includes the baselength of the rangefinder AND the length of roller's arm) ; also I checked (at sight) the cam of some lenses... the old Elmar seems indeed to have a "flat" cam (see Giordano post about this) while the Summilux 50 (old) looks to have a shaped cam : my only conclusion, and having not so much time to dedicate to this intriguing question , is that the answer is not so easy... : I suspect (JUST a suspect, having seen the math) they followed a classic engineering approach : 1) a "general formula" gives a profile too complex to machine 2) an "approximate formula" (which has, at its limit, the "flat" cam of the Elmar 50) does need some adjustement for lenses with critical DOF (50 1,4 and similar, and long focals) 3) Let's translate those adjustment in modifications of the theorical simple profile of the cam, modifications that are sufficiently easy to machine (I say this because, looking at the cam of Summilux 50, it seems to me that the cam profile has not a constant gradient, but a sort of "smooth transition", rather visible, form a profile to another) To be clear : I do not pretend to be right... problem isn't trivial and I scratched only the surface of it. 2 Link to post Share on other sites More sharing options...
120 Posted May 13, 2013 Author Share #18 Posted May 13, 2013 [article] "Although it would be possible to calculate what the profile should look like, I think it is just as easy to work out the correct profile by trial and error." 1 Link to post Share on other sites More sharing options...
Recommended Posts
Create an account or sign in to comment
You need to be a member in order to leave a comment
Create an account
Sign up for a new account in our community. It's easy!
Register a new accountSign in
Already have an account? Sign in here.
Sign In Now