Bizarre SR problem: Sensor shaking by itself

bwDraco · 03-25-2012, 06:15 PM

Just found this YouTube video:

You Tube

I've never heard of a problem of this nature. It's probably not something to be worried about, since this is likely to be an isolated failure. Any idea why the SR system would do this?

--DragonLord

**jbinpg** · 03-25-2012, 07:00 PM

As it happens, I think I know what the problem is. When I had finished the wiring on my battery booster power supply, I hooked it up to my K-5 after doing an initial adjustment on the voltage regulator pot to see how it would work. Well, the sensor shift started to go nuts - vibrating uncontrollably. I measured the voltage again and it turns out I had screwed up the first reading. It was only putting out 6.5 instead of 8.3 VDC. So I adjusted the regulator pot up, tried again and everything was fine. So my guess would be that the power supplied to the camera is too low.

Jack

**Digitalis** · 03-25-2012, 07:36 PM

I have had this problem with my K7 - especially with long exposures with long exp NR enabled - when the camera takes the dark frame you can hear the sensor dancing the rumba in there**. My trusty old K10D seems to also developing a shaky sensor - even exposures of 1/800th with the FA31 can show camera shake - sometimes this even happens with the IS turned off* It appears the pentax built in IS system is extremely sensitive to voltage fluctuations

*this happens only with bulb exposures on my K-7, though I'm sure it also occurs with the K-5.

**and I have tested this with the camera mounted on a sturdy tripod. though the battery was at less than full charge.

RioRico · 03-25-2012, 08:38 PM

I put a fisheye screw-on onto a midrange zoom mounted on my K20D. I put the camera+lens down on a table and power-up the camera, switching-on SR and LiveView. The camera+lens hop around on the table. The moral of the story: SR is stronger than you imagine.

DeadJohn · 03-25-2012, 09:49 PM

If someone were to pick that camera up off the table I think the SR would calm down. The video demonstrates why SR should be turned off when on a tripod or similar mounting. I can get the same sort of oscillation with my K-r on a tripod, on a table, or connected to a telescope. It never happens handheld.

(granted, the camera in that video might be broken and still wobble handheld, but normal SR behavior on a working camera has problems when not handheld.)

Philoslothical · 03-25-2012, 11:31 PM

It does look like it only started when he gave it a slight shake and set it down on the table. I'll have to see if I can induce this behaviour in my K-r, too, just out of curiousity.

This could turn out to be a perfect example of what happens when someone is unfamiliar with a system and goes looking for problems.

bwDraco · 03-26-2012, 07:06 AM

I'm suspecting the accelerometer might be broken...

--DragonLord

DeadJohn · 03-26-2012, 09:39 AM

There are many other threads that recommend turning off SR on a tripod. Any rigid surface, such as a table or the top of a stone wall, acts similarly to a tripod.

The camera might be broken, but that video doesn't show proof. Turn off SR and the camera probably works fine on a table. Handheld, the camera probably works fine whether SR is on or off. This is an odd quirk with Pentax' in-body shake reduction. It's not a Pentax weakness - neither Canon nor Nikon have in-body shake reduction.

**jbinpg** · 03-26-2012, 11:00 AM

I had my O-GPS1 unit active in astrotracer mode when it happened. This needs bulb mode by default and the astrotracer function by design shifts the sensor. And you need a tripod to do any serious astrotracer work. So the only thing that makes sense to me is that the power draw was too high for the given voltage and the sensor stabilization by the electromagnetic field failed. I reurned the voltage to nominal and everything began to work to spec.

Jack

bwDraco · 04-06-2012, 07:18 AM

I'm running into this problem to a smaller degree with my replacement K-5 (see https://www.pentaxforums.com/forums/pentax-k-5-forum/180910-k-5-being-exchanged-spots-accumulating-too-quickly-sensor-dust-beneath-oplf.html): the sensor is not moving freely on the image plane. Using the composition adjustment function, moving the sensor to the extreme top or bottom may intermittently trigger this behavior as the SR system is trying to get the sensor to the correct position. Note, though, that the sensor is not shaking as heavily as shown in the video above, and the problem does not occur when the sensor is centered. I'm at a loss as to why the sensor is getting stuck. If this problem doesn't clear up over the next few weeks, it's being exchanged again.

--DragonLord

**fs999** · 04-06-2012, 07:35 AM - **1 Like**

but now it is dust free

**DAZ** · 04-06-2012, 10:06 AM

It looks like a good example of resonates feedback. When the camera is in your hand there is no feedback just random input. On a tripod when the sensor moves the tripod can move a little. When the sensor moves back the tripod moves but not at the same time. The accelerometers detect the camera movement and the sensor is commanded to move in the opposite direction. With the moments out of synch a resonates can start. From the video you can see he starts the camera moving to make an input to start the whole thing. Any system with a feedback system can do this unless there is something in the system to stop any resonates.

DAZ

bwDraco · 04-06-2012, 03:04 PM

No, it's not a feedback loop causing the problem. Testing the SR system using the composition adjustment feature shows that the sensor is intermittently getting stuck at certain points near the top and bottom of the sensor movement range.

--DragonLord

bwDraco · 04-07-2012, 11:35 AM

Further testing clearly shows that the sensor is getting stuck and does not appear to be seated properly. The problem is most apparent when the composition adjustment is set to adjust the image composition rightward (sensor moves left when looking from the back of the camera; remember that the lens produces an inverted image) and the camera is oriented left-side up so that the SR system is moving against gravity (which should normally never cause a problem). The sensor began to shake in an attempt to move to the correct position which it could not. At times, the sensor stopped moving altogether and the SR system kept making a knocking noise, which meant that the sensor was locked up altogether. Occasionally, when trying to move the sensor a certain direction using composition adjustment, the sensor tilted instead of moving linearly as normally expected. Furthermore, the Dust Reduction function sometimes does not work (no sound from the DR II system is audible when the camera, only the aperture motor resetting; and selecting Dust Reduction from the menu exits immediately instead of taking about a second to complete). However, softly hitting the base of the camera body with lens facing up appears to clear up the problem.

From this information, it is clear that the sensor is not seated properly on the image plane, causing the SR mechanism to get stuck trying to shift the sensor to the correct position during operation. I will exchange the camera once B&H reopens.

--DragonLord

bats · 04-07-2012, 05:32 PM

Happens to my K5 occasionally. I wrote about it in another thread about 6 months ago. (https://www.pentaxforums.com/forums/pentax-k-5-forum/154046-sensor-flop.html)

I get 5-6 slaps per second.

In my case, the trigger seems to be 1) Live View and 2) portrait orientation of the camera.

It doesn't happen often. I live with it. The cure seems to be to power off the camera for awhile, then power back up.

04-06-2012, 07:18 AM	#10
bwDraco Veteran Member Join Date: Jul 2010 Location: New York Photos: Albums Posts: 1,071 Original Poster	The sensor is not moving freely I'm running into this problem to a smaller degree with my replacement K-5 (see https://www.pentaxforums.com/forums/pentax-k-5-forum/180910-k-5-being-exchanged-spots-accumulating-too-quickly-sensor-dust-beneath-oplf.html): the sensor is not moving freely on the image plane. Using the composition adjustment function, moving the sensor to the extreme top or bottom may intermittently trigger this behavior as the SR system is trying to get the sensor to the correct position. Note, though, that the sensor is not shaking as heavily as shown in the video above, and the problem does not occur when the sensor is centered. I'm at a loss as to why the sensor is getting stuck. If this problem doesn't clear up over the next few weeks, it's being exchanged again. --DragonLord Last edited by bwDraco; 04-06-2012 at 08:00 AM.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Live view problem with K20d ? shaking a lot!	TanGU	Pentax DSLR Discussion	13	05-21-2017 12:49 PM
Help! Bizarre problem with my beloved K5!	hsu1	Pentax K-5 & K-5 II	7	12-26-2011 06:40 AM
Shaking problem in liveview on a pentax k-x	commanderWhatNow	Troubleshooting and Beginner Help	8	05-29-2011 07:20 AM
k10d sensor shaking like mad	occhicone98	Pentax SLR Lens Discussion	2	12-28-2008 08:58 AM
HELP PLEASE !!! Bizarre K10d exposure problem!!	StephenG	Pentax DSLR Discussion	15	10-02-2008 12:52 PM

03-25-2012, 07:00 PM	#2
jbinpg Site Supporter Join Date: Mar 2008 Location: Prince George, BC Photos: Gallery \| Albums Posts: 3,546	As it happens, I think I know what the problem is. When I had finished the wiring on my battery booster power supply, I hooked it up to my K-5 after doing an initial adjustment on the voltage regulator pot to see how it would work. Well, the sensor shift started to go nuts - vibrating uncontrollably. I measured the voltage again and it turns out I had screwed up the first reading. It was only putting out 6.5 instead of 8.3 VDC. So I adjusted the regulator pot up, tried again and everything was fine. So my guess would be that the power supplied to the camera is too low. Jack

03-25-2012, 07:36 PM	#3
Digitalis Site Supporter Join Date: Mar 2009 Location: Melbourne, Victoria Photos: Gallery Posts: 11,694	I have had this problem with my K7 - especially with long exposures with long exp NR enabled - when the camera takes the dark frame you can hear the sensor dancing the rumba in there*. My trusty old K10D seems to also developing a shaky sensor - even exposures of 1/800th with the FA31 can show camera shake - sometimes this even happens with the IS turned off It appears the pentax built in IS system is extremely sensitive to voltage fluctuations this happens only with bulb exposures on my K-7, though I'm sure it also occurs with the K-5. *and I have tested this with the camera mounted on a sturdy tripod. though the battery was at less than full charge.

03-25-2012, 08:38 PM	#4
RioRico Veteran Member Join Date: Jun 2008 Location: Limbo, California Posts: 11,263	I put a fisheye screw-on onto a midrange zoom mounted on my K20D. I put the camera+lens down on a table and power-up the camera, switching-on SR and LiveView. The camera+lens hop around on the table. The moral of the story: SR is stronger than you imagine.

03-25-2012, 09:49 PM	#5
DeadJohn Pentaxian Join Date: Jan 2011 Location: New York Posts: 4,834	If someone were to pick that camera up off the table I think the SR would calm down. The video demonstrates why SR should be turned off when on a tripod or similar mounting. I can get the same sort of oscillation with my K-r on a tripod, on a table, or connected to a telescope. It never happens handheld. (granted, the camera in that video might be broken and still wobble handheld, but normal SR behavior on a working camera has problems when not handheld.)

03-25-2012, 11:31 PM	#6
Philoslothical Veteran Member Join Date: May 2011 Posts: 1,889	It does look like it only started when he gave it a slight shake and set it down on the table. I'll have to see if I can induce this behaviour in my K-r, too, just out of curiousity. This could turn out to be a perfect example of what happens when someone is unfamiliar with a system and goes looking for problems.

03-26-2012, 07:06 AM	#7
bwDraco Veteran Member Join Date: Jul 2010 Location: New York Photos: Albums Posts: 1,071 Original Poster	I'm suspecting the accelerometer might be broken... --DragonLord

03-26-2012, 09:39 AM	#8
DeadJohn Pentaxian Join Date: Jan 2011 Location: New York Posts: 4,834	There are many other threads that recommend turning off SR on a tripod. Any rigid surface, such as a table or the top of a stone wall, acts similarly to a tripod. The camera might be broken, but that video doesn't show proof. Turn off SR and the camera probably works fine on a table. Handheld, the camera probably works fine whether SR is on or off. This is an odd quirk with Pentax' in-body shake reduction. It's not a Pentax weakness - neither Canon nor Nikon have in-body shake reduction.

03-26-2012, 11:00 AM	#9
jbinpg Site Supporter Join Date: Mar 2008 Location: Prince George, BC Photos: Gallery \| Albums Posts: 3,546	I had my O-GPS1 unit active in astrotracer mode when it happened. This needs bulb mode by default and the astrotracer function by design shifts the sensor. And you need a tripod to do any serious astrotracer work. So the only thing that makes sense to me is that the power draw was too high for the given voltage and the sensor stabilization by the electromagnetic field failed. I reurned the voltage to nominal and everything began to work to spec. Jack

04-06-2012, 07:35 AM - 1 Like	#11
fs999 Loyal Site Supporter Join Date: Jul 2008 Location: Luxembourg Photos: Gallery Posts: 8,639	but now it is dust free

04-06-2012, 10:06 AM	#12
DAZ Loyal Site Supporter Join Date: Sep 2006 Location: Everett, WA USA Photos: Gallery Posts: 744	It looks like a good example of resonates feedback. When the camera is in your hand there is no feedback just random input. On a tripod when the sensor moves the tripod can move a little. When the sensor moves back the tripod moves but not at the same time. The accelerometers detect the camera movement and the sensor is commanded to move in the opposite direction. With the moments out of synch a resonates can start. From the video you can see he starts the camera moving to make an input to start the whole thing. Any system with a feedback system can do this unless there is something in the system to stop any resonates. DAZ

04-06-2012, 03:04 PM	#13
bwDraco Veteran Member Join Date: Jul 2010 Location: New York Photos: Albums Posts: 1,071 Original Poster	No, it's not a feedback loop causing the problem. Testing the SR system using the composition adjustment feature shows that the sensor is intermittently getting stuck at certain points near the top and bottom of the sensor movement range. --DragonLord

04-07-2012, 11:35 AM	#14
bwDraco Veteran Member Join Date: Jul 2010 Location: New York Photos: Albums Posts: 1,071 Original Poster	Further testing clearly shows that the sensor is getting stuck and does not appear to be seated properly. The problem is most apparent when the composition adjustment is set to adjust the image composition rightward (sensor moves left when looking from the back of the camera; remember that the lens produces an inverted image) and the camera is oriented left-side up so that the SR system is moving against gravity (which should normally never cause a problem). The sensor began to shake in an attempt to move to the correct position which it could not. At times, the sensor stopped moving altogether and the SR system kept making a knocking noise, which meant that the sensor was locked up altogether. Occasionally, when trying to move the sensor a certain direction using composition adjustment, the sensor tilted instead of moving linearly as normally expected. Furthermore, the Dust Reduction function sometimes does not work (no sound from the DR II system is audible when the camera, only the aperture motor resetting; and selecting Dust Reduction from the menu exits immediately instead of taking about a second to complete). However, softly hitting the base of the camera body with lens facing up appears to clear up the problem. From this information, it is clear that the sensor is not seated properly on the image plane, causing the SR mechanism to get stuck trying to shift the sensor to the correct position during operation. I will exchange the camera once B&H reopens. --DragonLord

04-07-2012, 05:32 PM	#15
bats Senior Member Join Date: Sep 2010 Location: San Francisco Photos: Gallery Posts: 102	Happens to my K5 occasionally. I wrote about it in another thread about 6 months ago. (https://www.pentaxforums.com/forums/pentax-k-5-forum/154046-sensor-flop.html) I get 5-6 slaps per second. In my case, the trigger seems to be 1) Live View and 2) portrait orientation of the camera. It doesn't happen often. I live with it. The cure seems to be to power off the camera for awhile, then power back up.