I have been testing my adaptall PKA mounts.
I was slightly surprised, once I had rooted out the ones clandestinely lurking on a lens or put absent-mindedly to one side, to see that I have 9 of them. I had previously marked 4 as iffy or faulty. My test was pretty straightforward; test/check mechanical function by inspection, then function, and exposure results in Av, on camera. I used a 52B f2.5 90mm macro, and a 01B f2.5 24mm. I was using the whitewashed wall in the conservatory as a relatively uniform exposure board, although the 24mm was including the windows in the edge of the shot. The results were as follows (pentax K5, centre weighted metering, 2M jpg's):
- All except two functioned fully. One didn't get the camera to display f numbers, no matter how much wiggling of the lens on the camera (usual fix) I did, and then I realised it was the one with the iffy bent aperture lever that I got stuck on the camera previously and posted about - took that off the camera quick!! In fact the small auto button on the side of the mount didn't click when setting the mount to f32/AE - the usual reason for that is that one of the "tabs" that engage with the aperture ring on the lens is out of position and I have noticed before that in this case Av doesn't function. The other wouldn't show f numbers below f4 - I think this is the one I had to disassemble and tlc on acquisition because it was completely seized up. I have to admit I was not perfectly diligent in making sure the delicate contacts that move over the circuit board (see pic here) didn't get nudged/tweaked.
- All registered f2.4 with these f2.5 lenses. One or two stuck at f2.8 initially, corrected with a varying amount of wiggling.
- All showed a significant underexposure (no compensation on camera) wide open. This was consistently around 0.5-1 stop.
- Then stopping down exposures would tend to brighten. At f11-ish exposures were typically good, or even tending to overexpose a little.
- However some mounts showed considerably more variation than others. A couple were almost consistent, with only slight change.
- And I felt I was seeing more exposure change with the 24mm.
These "filmstrips" illustrate one of the most consistent result sets (top) with the 52B, and one of the least (below) with both the 52B and 01B.
Mount 1 - 52B
Mount 2 - 52B
01B. You can see there is a shift in the histogram from f2.8 to f16 to the right of at least a stop.
For me the most annoying thing to occur when using these is pronounced exposure inconsistency. One twiddles the dial to stop down one or two stops, fires a shot and the image is overexposed. Is it the mount? Is it the camera? Is it the lighting situation?
I think the inherent underexposure, and a certain trend for that to alleviate with higher f stops, can be attributable to the camera. It is observable with other non-chipped PKA mount lenses, this filmstrip shows the results I got with my SMC-A 50mm f1.7 for comparison.
An element of underexposure is there but otherwise pretty consistent actually. T-mount and M42 lenses show both underexposure and a trend to higher exposure with higher f stops. But with those lenses the camera is stop-down-metering ie measuring the exposure to produce the shutter speed at the actual f-stop. My understanding is that with PKA lenses the camera calculates the shutter speed from the wide open aperture reading. It is noticeable that the shutter speeds mostly change correctly with the f-stops, halving each time. It is also interesting that the most noticeable inconsistencies are through f2.4-5.6. I was contemplating the possibility that there could be a mechanical issue, that some mounts are not fully effective at allowing the aperture to stop down fully (see my post #2843 ). But in that case I would expect the deviation from consistent exposure to be most noticeable at the higher f numbers (is that what we are seeing in the case of the 01B filmstrip - the shutter speeds are consistent but the exposure still moves - by more than a stop, look at the histogram?), and that the exposure might not correlate with the shutter speed. It's as if the actual stop the lens goes to is not the same as what the camera calculates. This might imply that the initial discrepancy in the f stop, f2.4 registered, f2.5 actual, is a factor. But my understanding of that is that this shouldn't be a problem. The initial f stop is slightly out, and so are all the rest by the same amount, it's a consistent error. Am I wrong?
I think I should identify the ones that showed the most exposure variation and check them out some more.