Originally posted by Loren E Is this specific to the 50-135 since it internally focuses, versus the extending 16-50 and 60-250?
I think this particular flaw is specific to the 50-135mm, but I don't know because I've never used any other SDM lenses. When you turn the focus ring on the 50-135mm, once you reach either limit (close focus or infinity), you can continue to turn the focus ring past that limit. The mechanism makes a "clunk" sound as you reach the limit, but the ring can keep turning, albeit with a small amount of added resistance. Doing this means you're turning against the internal clutch, which wears out that clutch over time. Eventually, the clutch no longer provides enough friction to turn the focus mechanism, especially as the lens is used in colder temperatures.
None of this explains the SDM failure of course. I think the clutch wearing out is another symptom of the underlying problem. The focus mechanism requires too much force to operate. So the SDM motor is overworked most of the time, especially when the lens is cold. I saw the first symptoms of failure at temperatures of 40f, about 9 months after I bought the lens. When temperatures warmed up, the SDM motor started acting "normally" again.
Another thing that anyone can try: the screw-drive mechanism. If you try to turn the screw-drive by hand, you'll find that it takes 17 turns to go from infinity to close focus. SEVENTEEN!!! Most Pentax lenses need 3 to 7 turns at most. Some macro lenses need 10 turns. It seems the Pentax engineers were aware of the fact that the focus mechanism requires excessive force to move, and so they down-geared it excessively to reduce the required torque. But I guess they didn't down-gear it enough, because the SDM motor still gets overwhelmed by it.
(this would also explain why the 50-135mm is one of the slowest-focusing lenses in the Pentax lineup)