[jaeaetee]

Hi!

I did some searching about the live view sensor shake phenomenon in some of the Pentax models. Most of the threads seemed to end to conclusion that the camera body is faulty and needs repair. I have a couple of yrs old K-7 that had never done that.

However, I recently bought a Cosina manual lens (24mm 2.8), that didn't stop down using the green button, because of the anodized aluminium bayonet that insulates all electrical contact pins. So I sanded some of the anodizing off the bayonet. Now everything works very well, but in live view the sensor shakes wildly. If I clean the contacts carefully and put the lens on a couple of times, live view might work normally for a short time, then the sensor starts shaking again.

No other lens does this (DA 35mm, 55mm SDM, Tokina 60-120 f/2.8, Tamron 24mm adaptall f/2.5, kit lens, Tamron zoom.. ).

Any ideas, what could cause this..?

[manacho2005]

It sounds like the camera is reading the lens as a long telephoto.
You may need to affix some foil to the lens mount or replace the lens mount with a metal one.

[Stone G.]

HI, I have the Cosina Cosinon-W 35mm/f2.8 that has undergone exactly the same modification. All I can say is that I have never had similar problems with that lens on my K-5 - and I've just tried it repeatedly after reading your post.

One long shot, presumably totally irrelevant: I have noticed that the aluminium (I think it is) base oxidizes over time and may require a renwed sanding once in a while. Perhaps(????) intermittent contact failure may fool or confuse your camera as to what lens is mounted.

As said, this is a looooong shot, but you may want to give it a try?

[Na Horuk]

What focal length did you input? You might want to try mimicking the contacts of an A lens of the same focal length and speed

[demp10]

Do you by any chance have the camera mounted on a tripod? If yes, you must turn SR off, otherwise the camera will vibrate significantly. Same if you place the camera on a solid surface like a table.

It will be worse if the camera thinks it has a telephoto.

[jaeaetee]

HI, I have the Cosina Cosinon-W 35mm/f2.8 that has undergone exactly the same modification. All I can say is that I have never had similar problems with that lens on my K-5 - and I've just tried it repeatedly after reading your post.

One long shot, presumably totally irrelevant: I have noticed that the aluminium (I think it is) base oxidizes over time and may require a renwed sanding once in a while. Perhaps(????) intermittent contact failure may fool or confuse your camera as to what lens is mounted.

As said, this is a looooong shot, but you may want to give it a try?

Well that could be it. I made sure that I sanded away all of the anodized layer (if too little sanding, only the black color is removed), but now when you said that it actually seems like a logical reason that some insulation layer develops onto the aluminium very quickly. I'll check what happens with some foil tape.

Last night I shot some handheld images with 1/4...1/10 sec shutter times with (almost) no blur, so the SR seems working fine, just the live view image is bouncing here and there. And yes, I've input 24mm as the FL.

[MSL]

Well that could be it. I made sure that I sanded away all of the anodized layer (if too little sanding, only the black color is removed), but now when you said that it actually seems like a logical reason that some insulation layer develops onto the aluminium very quickly. I'll check what happens with some foil tape.

I'm doubting that is the problem for two reasons. Aluminum oxides incredibly fast. Second - when you put a piece of foil tape in there, what is the tape made of - aluminum I am guessing, and it will be fully oxidized on both surfaces. However, the oxide layer is really thin and won't interfere with the conductivity.

[jaeaetee]

Found out what the problem was;

The entire lens should be earthed/connected to the camera body. I noticed that switching af/mf mode made a difference sometimes. This had something to do with the AF motor shaft taking a contact to one of the screw heads in the Cosina lens mount.

I sanded more anodizing off, this time behind the lens bayonet flange and around the AF motor shaft location, so that it makes a good contact to the camera body's mount spring --> no more sensor bouncing

Seems that the non-earthed lens formed some kind of (perhaps capacitive) interference circuit with the sensitive high-speed circuits of the SR system. Maybe Pentax is sensitive to this kind of interference..?

Sanding down only the flat part of lens bayonet didn't probably help because the insulating anodized layer is a bit thicker / higher than the sanded part --> no connection to the body mount.

[jaeaetee]

http://www.aohc.it/pshw2003/23fa3580_faj1835.jpg

BTW, there seems to be a wide contact in plastic bayonet Pentax lenses, just to do that earthing to the body mount..

[Stone G.]

Great jaeaetee! You solved your problem - and gave the rest of us Cosina users some valuable information too.

[jaeaetee]

Great jaeaetee! You solved your problem - and gave the rest of us Cosina users some valuable information too.

Thanks! Here's a photo of the bayonet.. for anyone not wanting to use alumium tape, I recommend sanding under one of the "ears" of the bayonet (in addition to sanding the data pin location), cannot be seen in the photo..