Replies to This Discussion

Permalink Reply by abey on May 23, 2009 at 2:41am

We tried two lenses: Philips CAY046 and CAX100, f = 4.6 and 10mm, respectively. They have a FOV between 3 and 6°, so fairly narrow by anyone's standard.

• ▶ Reply

Permalink Reply by Harry on May 26, 2009 at 8:04pm

Thanks, is a collimating lens also telecentric?
Also, do you know how to determine the image size shown in this USAF test chart?

• ▶ Reply

Permalink Reply by abey on May 27, 2009 at 2:33am

I'm not sure how to answer your questions. For the lens, you'd have to check the datasheet. As for the USAF test chart, I'm not at all familiar with it.

• ▶ Reply

Permalink Reply by automatik on May 20, 2009 at 11:31pm

@abey - interesting video. I also "poked" around your research site - you guys are doing some interesting stuff. let us know when you PhD thesis becomes available online

• ▶ Reply

Permalink Reply by Harry on May 26, 2009 at 11:55pm

Avago lists the "field coverage radius" as 1 mm. The datasheet shows this diagram of the lens and the sensor, should I interpret the 1 mm radius as 1/2 of the size of the 16x16 pixel array?

• ▶ Reply

Permalink Reply by Harry on May 28, 2009 at 10:17pm

abey,
No more questions on optics from me. I found an achromatic lens to do what I want and when used with an aperature at the focus, you get a telecentric lens. Telecentric lenses dont show perspective.

I've since experimented with a different approach that uses the stock lens of the mouse. The purpose of all this was to use a mouse chip for stabilization. The experiment involved a clear bottle of soap. I shook the soap up good to get a froth. As the bottle was tilted, the mouse cursor moved. There's a bias, because of the setup, that causes the cursor to want to always go in one direction easier than another. I dont fully understand that, but the mouse read the bubbles.

The next step is to find a liquid that reacts to pitch and roll and is something the mouse chip can see. The bubbles are something the chip can see, while the mouse cant see a uniform liquid. This wont be easy because:

1. visocsity will be a factor
2. bubbles rise to surface
3. the vessel is a factor
4. I'm not a physicist or programmer

• ▶ Reply

Permalink Reply by evilmacaw on May 29, 2009 at 8:11am

Harry,
you ever see one of the oil/water wave machines ? rectangular clear tube with blue oil and water that rocks end to end making wave forms in the tube .

• ▶ Reply

Permalink Reply by Harry on May 29, 2009 at 11:35am

Lava lamp oil of different colors, or even olive oil and basalmic vinegar maybe. It needs a flat clear surface that's for sure. The backwash of one of those wave devices could throw it off. Some kind of micro ball that floats might do it. As the vessel is tilted, the micro balls would seek the surface and go flying past the sensor.

Know where I can get flat clear tubes?

• ▶ Reply

Permalink Reply by Harry on May 29, 2009 at 5:49pm

Well, I think I understand the "bias" I saw with the soap bubbles and that is gravity. As the bubbles and the soap slide down the sides of the bottle(sensor over air space), the motion is detected. Gravity is pretty constant, so that's why the cursor always drifts toward one area. The alignment of the x,y of the chip determines the indicated bias.

I'm not so sure this will work as a pitch and roll sensor, but it might be an accelerometer.

• ▶ Reply

Permalink Reply by Harry on May 29, 2009 at 9:28pm

Geeeeeezuz, I did some reading and accelerometers Are used to indicate tilt. I've got a new idea for using the bubbles. There is some sort of gyro that uses a standing wave. As that sensor tilts, the standing wave tilts at a different angle. If I had a steady stream of bubbles from whatever that the mouse sensor could see, then a jolt of acceleration could alter the direction of bubbles. If this jolt is enough for the sensor to see, I'll have to investigate. Messing with the soap bottle has shown quite a bit so far.

If you have any comment or suggestions, fire away.

• ▶ Reply

Permalink Reply by Harry on June 17, 2009 at 12:37am

I havent given up on this, just yet. The current idea is to attach sensors(FETS) to the wing tips and the difference in voltage, due to the gradient of the atmospheric ionization, will indicate attitude. Maynard Hill did it first using a piece of radioactive polonium. My hunch is that the DC electronics at the time were difficult and FETS werent around. This voltage difference could move an electroactive polymer actuator in X-Y. The mouse sensor would read the object surface driven by the actuator.

This link is to a company that makes electroactive polymer devices:
http://www.artificialmuscle.com/products/P-25.html

Where can I get one of these?

• ▶ Reply

• ‹ Previous
• 1
• 2
• 3
• Next ›
• Page