Sudija

Hi everyone,
I have a little awkward question here but I am sure someone who knows a little bit more about electronics can give me good advice.

I took apart the lens from a consumer video camera and took out the Optical Image Stabilization module in an effort to figure out how it works and possibly hijack the signal so I can control the movement.

In the camera this module is getting the signal from the accelerometer so it knows how to engage the two electromagnets and counteract the movement (shaking).

What I have done is take the leads that go into it and try to fake what gyro sends to this element. So far I have noticed that when I short 1 and 2 or 3 and 4 the element goes to one side or up. So I put a potentiometer on it (2K ohm) and it gives me some control by turning the dial, but not good enough.

Has anyone any idea how can I achieve a gradual movement of this lens element vertically and sideways with a slider or dial of some sort?

Your help would be greatly appreciated!

DonaldG

Hmmmm - A very interesting project.

You may have already thought of these ideas. I know less about anti-shake electronics than you do but, thinking out loud:

I would assume that the 'at rest', 'no shake' position of the elements must be the centre of the individual element travel distances.

what is the driving motor? is it a servo?

What is the signal that is being connected/shorted? It is probably be at CMOS level (8v) rather than TTL level (5v)...maybe?

What are those 2 4pin devices? are they marked, if so Google it and get the schematic or device details... are they part of the equation? Are they bridge rectifiers?

You have tried and analogue (potentiometer) control... is it more suited to digital control using something like a 555 timer (or similar) to generate variable pulse widths. You would need to experiment with frequency range.

Thinking about the physical movement of the elements during a camera shake, it needs to track the speed of the shake as well as the transit distance of the shake. The angle of the shake will be a ratio between the x & y axis.

Are any of the contacts that you connect a 'ground'? Or are they 'above/below' ground?

What is the make of the camera that it came from? It may be possible to get technical tedails from the manufacturers website (a long shot, but you never know!)


A few points to ponder on.... I wish you well with the project.

__________________

Submit New Articles

Done_Fishin

I thought anti-shake was supposed to capture elements so fast that it could store the picture in memory then freeze the image before taking another shot at the scene being viewed. That way it avoids shake by taking very fast single shots.

__________________

D_F

craigwatanabe

When you say not good enough, is the control too jerky or not enough movement?

There are two types of potentiometers, Linier and Audio taper. Linear taper is basically a straight line change in resistance as the pot is rotated. Audio taper allows for finer incremental changes at either end but greater change in the middle.

In any case it seems shorting any of those leads results in a predetermined movement to a particular side. A potentiometer won't graduate that change slowly it will just put a variable resistive path to closing a circuit in this case an electromagnet to pull the lens one way or another. In other words the lens will stay put until there's enough voltage present to make the electromagnet powerful enough to pull the lens.

Unfortunately with the current setup there is no gradual control to be achieved for the purposes you describe.

__________________
A computer is only as powerful as it's user so build wisely.

DonaldG

Personally, I would believe them to be servo motors and would require digital drive rather than analogue

__________________

Submit New Articles