Grytpype
Well-known
It's a well known problem with the post-war Contax that the cement holding the plano-convex/plano-concave prisms in the rangefinder 'swinging wedge unit' shrinks and cracks with age, thus misaligning the RF image (later examples had a different cement, and don't have the same problem). There is no way to correct the vertical alignment except to re-cement the prisms. Having finally plucked up the courage to attempt this on a IIa, I thought I would add my contribution to previous threads on this subject. I won't go into a dismantling 'how-to'. There is not much to dismantle and this is covered well enough in the manual. The method I've described won't be necessary in all cases. It may be possible just to scrape out some of the loose cement, correct the alignment, and then cement in place. I could not see that this would be possible or reliable, at least on the wedge-unit I was using.
I spent some time trying to figure out what I actually needed to do. Did I need to raise/lower the fixed lens or tilt it? I came to the conclusion that, because the curved surfaces of the 'wedges' in the post-war Contax are lens-shaped, the only thing that affected the vertical position of the RF image was the vertical position of the axis of the lens. In other words tilting the element would only have an effect in so far as it affected the position of the axis. Obviously it is best to keep the flat surface of the lens square to its housing if possible, and just move it vertically. The pre-war Contax and Kiev is a different case entirely, because the 'wedges' are single-curvature: i.e. cylindrical section. As far as I can see, moving either element vertically will have no effect on the vertical position of the RF image. Here the element must be tilted. I'd also assume that rotating either element would rotate the RF image. In the post-war Contax, rotating the elements won't have this effect, so all you really have to bother about is the vertical position of the axis. If I have any, or all, of this wrong, someone please let me know!
The camera in question was a fairly late one with a 'Q' serial prefix, and the 'swinging wedge unit' had the later type of cement, so in fact the vertical alignment was OK. However the fixed prism was cracked in two (very visible in the finder) and the long 'pencil' prism was the later type with a platinum semi-silvered surface (according to Henry Scherer) which had a cloudy separation, making the viewfinder and the RF image poor and hazy. I decided to change both the swinging-wedge unit and the long prism.
The long prism came from a 'T' prefix black-dial IIa. The swinging-wedge unit came from an 'A' prefix black-dial IIIa. The cement had failed, and there was a significant vertical error in its original camera: a much larger error in its new home. It appears that the pencil prisms are not absolutely identical as far as the alignment of their optical surfaces are concerned, and each swinging-wedge unit will be set up to match the prism in its own particular camera. You could possibly change both together without affecting vertical alignment, but if the parts come from different cameras, you will definitely have to reset the wedge alignment.
An unsatisfactory aspect of this job is that if the cement has failed on the fixed element of your swinging-wedge unit, inevitably it will also have failed on the moving element. I would not know where to start if I wanted to re-cement both elements. Fortunately the swinging element is rather better 'cradled' by the cement than the fixed element, so it has probably moved less. I assumed it was likely to be in roughly the correct alignment, and just ran some liquid 'superglue' into the cracks in the cement to stabilise it.
I seem to remember seeing somewhere on Henry Scherer's site that it is possible for the prism elements to be fitted upside down. I have not been able to see any asymmetry, but just in case, I took care first, with the camera lying on its back, to take off the little cover from the front of the fixed element and make sure there was some way I could be sure which way up the glass was. In fact there were a couple of tiny chips in the 'South East' corner of the glass, which made a handy permanent indicator.
After taking the wedge unit off the camera you should examine the position of the fixed element carefully, measuring or estimating its position vertically and horizontally, relative to its frame, its depth below the milled area where the front cover sits, and its distance from the moving element behind. The fixed element is only cemented at its sides. The clearances at top and bottom are large, so it is not really a good plan to think of fixing the element by just applying cement here. However the clearances at the side are small, so you cannot hope to scrape out the old cement and clean the surfaces with the element in place. I started chipping out the cement and the lens fell out almost immediately.
Having cleaned the glass and the frame, the next task was to make means of holding the wedge in position while setting the alignment and holding it in exactly the same position after applying cement. The device I came up with is pictured below, made of copper wire about 1mm in diameter, and it is attached to the flat face of the wedge with 'Evostick', chosen because it is very easy to release with a couple of drops of naptha. It is attached near the top and bottom edges with a clear space in the middle so that the RF alignment was clear. The extension to the side where the wedge is attached is to allow a straight-edge (the end of a ruler) to be used to judge the position of the glass relative to the front surface of the housing. I had to mount the wire with screws a little longer than the normal wedge-unit screws.
I spent some time trying to figure out what I actually needed to do. Did I need to raise/lower the fixed lens or tilt it? I came to the conclusion that, because the curved surfaces of the 'wedges' in the post-war Contax are lens-shaped, the only thing that affected the vertical position of the RF image was the vertical position of the axis of the lens. In other words tilting the element would only have an effect in so far as it affected the position of the axis. Obviously it is best to keep the flat surface of the lens square to its housing if possible, and just move it vertically. The pre-war Contax and Kiev is a different case entirely, because the 'wedges' are single-curvature: i.e. cylindrical section. As far as I can see, moving either element vertically will have no effect on the vertical position of the RF image. Here the element must be tilted. I'd also assume that rotating either element would rotate the RF image. In the post-war Contax, rotating the elements won't have this effect, so all you really have to bother about is the vertical position of the axis. If I have any, or all, of this wrong, someone please let me know!
The camera in question was a fairly late one with a 'Q' serial prefix, and the 'swinging wedge unit' had the later type of cement, so in fact the vertical alignment was OK. However the fixed prism was cracked in two (very visible in the finder) and the long 'pencil' prism was the later type with a platinum semi-silvered surface (according to Henry Scherer) which had a cloudy separation, making the viewfinder and the RF image poor and hazy. I decided to change both the swinging-wedge unit and the long prism.
The long prism came from a 'T' prefix black-dial IIa. The swinging-wedge unit came from an 'A' prefix black-dial IIIa. The cement had failed, and there was a significant vertical error in its original camera: a much larger error in its new home. It appears that the pencil prisms are not absolutely identical as far as the alignment of their optical surfaces are concerned, and each swinging-wedge unit will be set up to match the prism in its own particular camera. You could possibly change both together without affecting vertical alignment, but if the parts come from different cameras, you will definitely have to reset the wedge alignment.
An unsatisfactory aspect of this job is that if the cement has failed on the fixed element of your swinging-wedge unit, inevitably it will also have failed on the moving element. I would not know where to start if I wanted to re-cement both elements. Fortunately the swinging element is rather better 'cradled' by the cement than the fixed element, so it has probably moved less. I assumed it was likely to be in roughly the correct alignment, and just ran some liquid 'superglue' into the cracks in the cement to stabilise it.
I seem to remember seeing somewhere on Henry Scherer's site that it is possible for the prism elements to be fitted upside down. I have not been able to see any asymmetry, but just in case, I took care first, with the camera lying on its back, to take off the little cover from the front of the fixed element and make sure there was some way I could be sure which way up the glass was. In fact there were a couple of tiny chips in the 'South East' corner of the glass, which made a handy permanent indicator.
After taking the wedge unit off the camera you should examine the position of the fixed element carefully, measuring or estimating its position vertically and horizontally, relative to its frame, its depth below the milled area where the front cover sits, and its distance from the moving element behind. The fixed element is only cemented at its sides. The clearances at top and bottom are large, so it is not really a good plan to think of fixing the element by just applying cement here. However the clearances at the side are small, so you cannot hope to scrape out the old cement and clean the surfaces with the element in place. I started chipping out the cement and the lens fell out almost immediately.
Having cleaned the glass and the frame, the next task was to make means of holding the wedge in position while setting the alignment and holding it in exactly the same position after applying cement. The device I came up with is pictured below, made of copper wire about 1mm in diameter, and it is attached to the flat face of the wedge with 'Evostick', chosen because it is very easy to release with a couple of drops of naptha. It is attached near the top and bottom edges with a clear space in the middle so that the RF alignment was clear. The extension to the side where the wedge is attached is to allow a straight-edge (the end of a ruler) to be used to judge the position of the glass relative to the front surface of the housing. I had to mount the wire with screws a little longer than the normal wedge-unit screws.
Before starting, make sure the horizontal RF adjustment is at about the middle of its range. It is controlled by a lock-screw and eccentric adjuster. I set up the element 'dry' initially, without cement. I could not count the number of times I got the alignment almost perfect, only to lose it completely immediately afterwards! Assembled dry, this was no problem. Assembled with cement it would probably have meant a complete strip and clean.
In aligning the element I attempted to place it as near as possible to the position noted previously, and "free floating", without interference from the frame. However the element in mine before it came out appeared to be as far as it could go towards the outer side of its frame, and I attempted to set it at this position when I aligned it. I think this was a mistake, and it contributed to a problem during final assembly which I will mention later. You don't need to worry too much about the horizontal alignment at this stage as long as it is 'ball-park' correct. I made sure the copper wire was right up against the fixing screws so that I could remount it at the same position after applying the cement. When I had the vertical alignment spot-on, I carefully took the element and wire off to apply cement.
Using a solvent based cement to fix the element is bound to be fraught with problems because it will inevitably shrink as it hardens and affect alignment. I decided to use an epoxy, because I was reasonably confident I could break the bond and clean the surfaces if I really had to, and also the curing time is long enough to give time to make sure everything is in correct alignment as it sets, and to bale out and start again if everything goes pear-shaped. I used J-B Weld, simply because it is dark grey in colour, so it looks appropriate.
I was careful not to apply too much of the epoxy, because I did not want it to squeeze out at the back and cement the fixed element to the moving element! After re-fitting the fixed element the vertical alignment was still correct, and remained that way as the epoxy hardened.
The problem I mentioned earlier raised its head when I came to refit the focus-wheel and its bushing. This bushing is quite a close fit around the fixed wedge. As I tightened it I realised that it was pressing on the 'South West' corner of the wedge, due to the fact the wedge was set lower than it had originally been, and also because I had set it as near to the outer edge as it could go. Rather than start the whole job again I filed a little from the bushing. Annoying. The 'mod' does not affect function, and is invisible from the outside, but I will always know it is there! I could have avoided this problem if I had done the alignment with the focus-wheel and bushing in position.
In aligning the element I attempted to place it as near as possible to the position noted previously, and "free floating", without interference from the frame. However the element in mine before it came out appeared to be as far as it could go towards the outer side of its frame, and I attempted to set it at this position when I aligned it. I think this was a mistake, and it contributed to a problem during final assembly which I will mention later. You don't need to worry too much about the horizontal alignment at this stage as long as it is 'ball-park' correct. I made sure the copper wire was right up against the fixing screws so that I could remount it at the same position after applying the cement. When I had the vertical alignment spot-on, I carefully took the element and wire off to apply cement.
Using a solvent based cement to fix the element is bound to be fraught with problems because it will inevitably shrink as it hardens and affect alignment. I decided to use an epoxy, because I was reasonably confident I could break the bond and clean the surfaces if I really had to, and also the curing time is long enough to give time to make sure everything is in correct alignment as it sets, and to bale out and start again if everything goes pear-shaped. I used J-B Weld, simply because it is dark grey in colour, so it looks appropriate.
I was careful not to apply too much of the epoxy, because I did not want it to squeeze out at the back and cement the fixed element to the moving element! After re-fitting the fixed element the vertical alignment was still correct, and remained that way as the epoxy hardened.
The problem I mentioned earlier raised its head when I came to refit the focus-wheel and its bushing. This bushing is quite a close fit around the fixed wedge. As I tightened it I realised that it was pressing on the 'South West' corner of the wedge, due to the fact the wedge was set lower than it had originally been, and also because I had set it as near to the outer edge as it could go. Rather than start the whole job again I filed a little from the bushing. Annoying. The 'mod' does not affect function, and is invisible from the outside, but I will always know it is there! I could have avoided this problem if I had done the alignment with the focus-wheel and bushing in position.
marcr1230
Well-known
This is spectacular!
Larry Cloetta
Veteran
Thanks for posting. Over and above the general camera expertise and bravery on exhibit here, I kept looking at the copper wire. Anyone who has ever tried to create a sequence of bends in one continuous wire while getting them in exactly the right position, and in exactly the necessary planes, will notice an abnormally high skill level here. You do this like an orthodontist. (A compliment in this case)
Hats off.
Personally, I'm not trying this
Hats off.
Personally, I'm not trying this
rajmohan-fotograf
rajmohan
Holy moly! Thank you - fascinating stuff
marcr1230
Well-known
One thing I don't follow is how you determined the adjustment, are you looking through the viewfinder as you move the wedge up/down?