Compur Rapid - getting the file out!

[Grytpype]

I've been cleaning the Compur Rapid shutter in an Ikonta 521/16 due to hesitant slow speeds. The Ikonta is in a remarkably good condition for an 80 year old camera, and there is absolutely no evidence that anyone has been inside before.

I refitted the retard unit at the marks I made before dismantling, and when I checked the shutter at the one second mark it gave an exposure of 995mS and I thought I was on a winner - until I checked the 1/2 sec time - 820mS! 1/5 sec was also very slow, and I didn't bother to go any further.

If you look at the images below you can see that at the one second setting, with the shutter cocked, the retard lever is not touching the speed-cam, but is resting on what the manual calls the "main lever" (I would rather call it the cocking-ring). Presumably, this must be wrong? If the retard-lever could get to the deepest part of the cut-out of the speed-cam it would have a longer stroke at 1 sec, which would make it slower, and if I then adjusted it correctly it would increase the other speeds as well.

The obvious answer is to file down the outside of the cocking-ring to make room for the retard-lever to move fully in the speed-cam. I am a little reluctant because previously, whenever I have decided that the only solution to a problem is to file something or bend something, I have realised afterwards what I actually should have done instead. This time I really think a file is the only answer, but does anyone have any other idea?

Steve.

​

 

[Mr_Flibble]

Personally, I would not touch it with a file, except as a very last resort.
And even then I don't think an 80 year old shutter should need that kind of tinkering.

My first choices would be to clean the slow speeds escapement and check the spring tensions of any torque or blade springs involved in it. Add a pin-prick of oil to the axels of all the gears in the mechanism.
Maybe add a bit of extra springiness with a slight bend or half turn to the springs.

I vaguely recall there's a coil spring in the slow speeds escapement that will run down if you remove it from the shutter. But I might be confusing that with the self-timer mechanism from the Compur-S

 

[Grytpype]

I've always considered the Compur a top-quality shutter, but my feeling is that it left the factory like this. It was set to be correct at 1 sec, and they did not bother too much about the other speeds.

The retard unit is cleaned and the pivots lubed with watch oil, and I know the spring of which you speak. All is well here!

You can see as it stands that it is physically impossible for the retard lever to run through its full stroke at 1 sec, as it surely should, due to the cocking ring overlapping the speed cam cut-away, and you can see how close (radially) the position of the retard lever is at 1/2 sec to its position at 1 sec. I don't think you could ever get both speeds to be close to correct.

At the present I am trying to find pictures of the cocking ring from other shutters to compare, and I do think this one seems rather wider in the relevant area

 

[Dan Daniel]

I always thought the basic speed setting for these shutters involved- first, set the overall escapement position for 1 second. Sounds like you have that nailed.

Then you bend that tab to give 1/10 second. This is the step it sounds like you need to do next.

Maybe I am off when it comes to the Compurs, but that's the standard procedure for Rapids and Synchros. I would try most anything before taking a file to the cocking ring itself.

 

[Mr_Flibble]

Ah, okay.

Hmm, I wonder, as most of this stuff was cast and machined, I would assume all the measurements would be well within the tolerances. It almost makes me think if something else is not installed correctly.
There is some playroom on how to attach the slow speed escapment to the shutter housing, influencing the movements and interaction with all the cams from the speed dial and cocking ring.

Or something is bent out of shape.

I spent weeks trying to figure out why B and T were not working on a Compur-S. Turned out a forked part had a tooth half a millimeter out of spec.

 

[johnnyrod]

I'm not sure why your looks like that, here is one I worked on, set at 1 second, and the pin has clearance to drop onto the speed setting ring:
1 second setting by John Rodriguez, on Flickr
Is yours cocked when you took the picture? Mine isn't. The main lever will have a different profile at that point but I don't want to open it up again to have a look! I do have another old Errtee with a similar shutter, it's not in use, i could have a look if you like.

 

[Grytpype]

Then you bend that tab to give 1/10 second. This is the step it sounds like you need to do next.

I think maybe that's just the later shutters. The riveted pin on this one is short and stubby, and not bendable.

I do have another old Errtee with a similar shutter, it's not in use, i could have a look if you like.

Thanks, but I think there seem to be too many different types of these shutters for easy comparison.

I've attached a picture of the part of the cocking-ring I would be filing (between the arrows). I cannot help noticing that all of the other outer surfaces of the part seem to be finished smooth, whereas this area is just 'as stamped' - rather unfinished in appearance! For what it's worth, I checked pictures of the shutter from an Ikonta 521/2 I stripped recently, and this whole area looks a nice uniform width.

 

[retinax]

That's suspicious indeed. But here's another idea: IIRC, on a Synchro Compur the position of the entire escapement needs to be adjusted. On the Rapid, too? Maybe that would be enough?

 

[Grytpype]

On the Synchro, as far as I remember you adjust the position of one end of the retard unit, and I think also you can bend the pin on the retard lever as Dan Daniel mentions.

On the Rapid (and plain Compur) you can adjust both ends of the retard unit, so you can adjust the depth of pallet engagement and also the amount of retard-lever movement, but neither will prevent the cocking ring restricting the movement of the retard lever, which is what is needed to increase the difference between 1 sec and 1/2 sec, I am sure.

I think I will go ahead with the filing. I don't like changing things that have been as they are for 80 years, but I am convinced it is wrong as it is. If I am wrong, it should only affect the 1 sec speed. If I am right, I will have the satisfaction of fixing a 80-year-old fault!

I may not have time over the next three days, but I will post results, for good or ill, as soon as I can.

 

[retinax]

If you're certain... I can't imagine negative consequences as long as the ring doesn't get bent or scratched in the process.

 

[Moto-Uno]

I feel your pain (really) , I'm into my Kowa 150mm shutter for the 4th time and these old mechanics hands are being tested by the delicate nature of everything inside 🙂
Good luck ! Peter

 

[Grytpype]

If you're certain... I can't imagine negative consequences as long as the ring doesn't get bent or scratched in the process.

I'll hold it at the 'handle' part clamped between pieces of plywood.

I found some pictures of the Rapid in an earlier 520/2, and, like the 521/2, the area clockwise of the part of the ring that controls the pallet-lever is a nice uniform narrow width (until it gets a little narrower just before the tab that drives the retard-lever). The cocking-ring in this 521/16 gets wider here by about .3mm.

 

[oftheherd]

Somewhere at home I have one of Ed Romney's books which among other things covers lens repair. I think it has the Compur dial set shutter in it. If I can find it I will look and see if it mentions anything that will help. One of his books used to be on line but I couldn't find it with Google so maybe his family had it taken down.

 

[Grytpype]

I said in my first post that whenever I decide to file or bend something, I realise afterwards what I should have done instead. In this case I should have done nothing!

The cocking-ring is hard-chrome plated, which cannot be filed, or abraded with abrasive paper, but does flake off. Also it turns out that it was not the cocking-ring that was blocking the retard-lever from moving all the way in the speed cam, but the aluminium barrel of the shutter.

I've positioned the retard unit to give the best results I can, and they are about as I would expect down to 1/10 sec (1/25 and 1/10 are very good). As before, 1/5 and 1/2 are half a stop slow but 1 second is accurate.

I was wrong about the peg on the retard lever not being bendable. On closer examination, it is waisted, and somewhat bent. It could possibly affect the slow-speed issue, though I am not going to risk bending it: break it and the camera's a write-off. The bendable waist is short and narrow, obviously meant to be set once and left, and not designed to be experimented with. It is not mentioned as an adjustment in the Compur Rapid manual, and presumably it was set at the factory

 

[farlymac]

Hope you didn't damage the cocking ring. Something like what you later found (out-of-round shutter barrel) can be really hard to diagnose. I watch the movements many times to get the full sequence down of what happens when, and even then it might take several dismantling cycles before hitting on what the problem is.

I had a doozy of a time with a Synchro-Rapid on a Contaflex 2 one time. The trigger block kept hanging up as it came around the outside of the shutter mech on what I thought was a high spot of the inner barrel (internal cocking through the film advance, so there is an inner and outer shutter barrel). I lubed (which is not my standard response, but I just had to try it). I filed. Nothing helped that block getting past the drag spot. But I kept watching the movements, and worked my way back through each part until I found a spring tail that had been over adjusted, which caused so much tension on the B lever the trigger block could not override it. And it was two levers away from the part that needed to move!

Yeah, it can be frustrating finding what is blocking a movement in the shutter, but it usually winds up being something that you'd never seen before, or is not very obvious.

PF