Originally posted by Lowell Goudge a long time ago I converted a ricoh 50mm F2 to PKA ... it seems that the aperture movement of K mount lenses cannot be successfully implemented unless the aperture lever is allowed to be fully released, which means that the only correct apertures are wide open and fully stopped down, all others have errors regardless of how you set them
Good information. You went further than I would to convert a lens. I am not sure with your description, could you confirm a couple points? Did you retry this lens on a current dSLR? And did you set the same aperture value both on the lens' aperture ring and within the camera body?
When I look at how the aperature works mechanically when off-camera, the aperture is sprung to completely close the aperture. The aperture ring moves not only the blades to the desired position but also the aperture lever to a coresponding position, the ring prevents the lens from closing further than a specific position. On an A lens, there is no difference in aperture between the smallest designed aperture and the A position. The only additional change is the A pin is extended from the lens.
When the lens is mounted to the camera, a cooresponding lever in the body pushes against the spring in the lens to fully open the aperture. On a non-A camera body during exposure, the body side of the lever would move to fully release the lens' lever - which would then stop at the pre-set point based on the aperture ring. However, on a A type camera body, when the camera believes an A type lens is attached, the body side of the lever moves only far enough to allow the lens to stop down to the aperture specified by the camera's electronics. The correct aperture is a percentage of the available lever movement, and why it is critical the body knows both the minimum and maximum available apertures.
In theory, if I have pre-shorted the A pin on the body and set the same aperture on both the lens and the camera, during exposure, the body side of the lever would stop movement at exactly the same distance as my lens' lever would allow. There are only two possible variations:
If the body side of the lever moves farther than the aperture set on the lens, the aperture would still be the value set on the lens because the lens would mechanically prevent the lens from stopping down further.
If the body side of the lever stops short of the aperture set on the lens, the aperture opening would also stop short of the value set on the aperture ring. The photograph would be overexposed according to the meter.
Unless the camera body has misinterpreted the lens' minimum and maximum aperture (and therefore how far the aperture lever would move for a given intermediate aperture), or the camera's meter misreads the exposure at full aperture, I cannot see how you could achieve under exposure. And overexposure whould be limited by however far short the body side lever travelled.
All of this assumes that a lens' lever movement correspondes to Pentax's calculations as to what percentage of movement equates to a specific f/stop. I think it is pretty safe to assume all lenses specificed as Pentax compatible do just this. I don't know if an alternate K-mount lens would do this. And I mean that literally, I don't know - I am not making any claim that alternate K-mount compatible lens do or don't.
What I can say is with my experimentation - and all my limited collection of lenses are all specified Pentax compatible - my hacking has produced exposure on my K-r body as accurate as what I see with my Pentax DA lens.
Given the equipment specified by the OP and how easy it is to make these easy, fully reversible modifications, it would be worth it to me to give it a try and see what happens.