Originally posted by willskywalker93 Just did this with mine after its second SDM bit the dust. Pretty annoying. I had noticed it being sluggish for a few months and needing to warm up. Left it sitting for a month for school, and it's dead. None of the usual wake up tricks did a thing. I want to go to the sigma because it has less flare/fringe/other aberrations than the 16-50, but I'll lose weather sealing and full time manual capabilities. What to do....
After an hour of trying to get to the debug menu of my K5 since I'd read that the K3 couldn't be used for lens rom edits, I gave up and switched the to KX. Found the MODSET.492 file name easily, took me forever to find a thread saying I had to hold the AV button down on boot to get me there. The 16-50 actually focuses much more accurately and quickly in screw drive now. The noise is just weird for my now, I've got the Sigma 70-200 HSM II, the DA 18-135 WR, and the BigmOS, so focus noise was a thing of the past when my 16-50 was functioning. Alas, it is back. Like an itch on that one part of your back that you just can't quite reach.
Anyway, rant over. Thanks for the guide, OP! Much appreciated. Good to have a working lens again.
I recently converted a 50-135 that was having focusing issues in low light in a k-1. My testing if the lens showed that while SDM was functional it seemed to fail to focus under some circumstances - perhaps giving up too soon as the SDM lagged perhaps. The conversion hasn't been tested by the user but I thought I would post this to see if others have seen this scenario - SDM works and isn't sluggish per say but fails to lock focus in poor lighting more than expected. In these cases has Conversion to screw drive helped? In this case on my k-3 and k100d super it seems to have had a positive impact. Waiting for customer to validate with his k-1.
Note: I tested my own 50-135 in the same focusing challenges and mine with SDM enabled did not fail nearly as often. 1/10 as often perhaps. The screw drive conversion was better than before perhaps failing 2/10 of the original 10 failure scenarios.