I have read through most of the thread, but it has been a while. I guess I should re-read it at some point.
Could also be lens, if something moves briefly and is stopped by software or mechanical limit.
That could be, but I was in manual focus, and I wasn't using a zoom script. Nothing was supposed to move at least. I wasn't watching the camera whenever it happened, so I don't know if anything moved. Unfortunately, I had the lens zoomed near 60mm, so if it went any farther I wouldn't notice it. It was still zoomed near the far limit when I found it. Whatever was clicking was extremely quiet. It was hard to hear over the static noise of the IS mechanism.
Camera RAM will be cleared as soon as the normal battery is removed. Most other camera non-volatile data is stored in onboard flash (the same "ROM" as the firwmare), which doesn't need any power at all. I don't think the CR1220 does much more than the clock, but I could be wrong. Certainly worth trying if anyone else gets this error.
That's what I would expect as well. I forgot about the CR1220 until I took the camera apart, otherwise I would have removed it for a while just as a test. I didn't notice anything being reset besides the clock. The file numbering sequence was pushed back a few hundred, but that probably has to do with the crash itself.
Some other thoughts:
- I assume your camera was already disassembled once, to remove the IR filter ? So in your case, there could be issues related to this (ribbon cabled not making good contact etc)
As an aside, if you want to post details of this process in http://chdk.setepontos.com/index.php?board=36.0 I'm sure other members would be interested.
I actually took the camera apart twice. I started out using a piece of a Lee 87 filter, then I switched to a Kodak Wratten 89B. I tried to be careful, but the parts are tiny so I certainly could have damaged something or had some dirt sneak in. I took pictures of the whole process with the intent of doing a writeup, I just haven't found time to do it yet. It is actually pretty simple.
The stock IR cut filter is a piece of glass 0.4mm thick. I couldn't find a source of filter glass that thin without ordering a truck load, so I just used gel. A small piece of gel filter cut to size fits fine. The stock filter is so thin that replacing it with gel doesn't change the optical system much. The manual focus scale is still close enough to correct that I would have to actually measure it to see any error.
- I wonder if the retract after a crash can leave the lens hardware in an unusual state, that isn't really broken but fools the software. There must limit switches, position sensors and so on. An unexpected combination of these might trigger the lens error without anything really being wrong. Or maybe there's a position where the actuators can stall or hit a current limit. In this kind of situation, just taking it apart and moving things around might clear it. The next logical step might be trying to gently manipulate the lens without taking it apart.
The weird thing is that the lens seemed to retract fine after the crash, then it refused to extend again. I sort of wonder if the code that handles startup after a crash has a bug and the camera hits some sort of double fault that isn't handled. I tried to force the lens in every direction while turning it on. I wasn't very gentle. I didn't expect to be able to fix the camera anyway, so I put quite a bit of force on it to see if I got lucky. Most of the times that I read about trying to force the lens it was after the camera had been dropped or some dirt got into it. My SX130 was never dropped and it is pretty clean except for some dust. My SX200 has had a much rougher life, and it still works fine.
I almost hope that the camera crashes again with the lens error. That way I can try things one at a time next time and maybe figure out the fix. Every time I have seen someone disassemble a compact camera lens, it never goes back together. I fully expected to be buying another camera later that night, so I wasn't very scientific.