PDA

View Full Version : Need programming help: video stabilizer



AMG AV
09-09-2008, 08:38 AM
I'm ordering the Servo Controller to experiment with some video mounts I use while capturing video from moving platforms including cars, boats, and aircraft. I've experimented with RC Aircraft gyros (accelerometers) to provide some stabilization of servo based gimbals. Parallax offers many sensors that may help to provide many levels of stabilization to my camera gimbal but I am not a programmer and would need help getting sensors to communicate with a servo controller. Need to have very fast response rates to outside influences (pitch,roll,azimuth) in order to counter-control camera gimbal. I am also not opposed to using other 12v servos outside the normal RC type servos. Not asking for a freebee here, just some advice and possible income opportunity for an experienced programmer. more details available for those who might be interested.

aes@mchsi.com (mailto:aes@mchsi.com)

DJSandberg
09-09-2008, 12:13 PM
When you say "Very fast response" how do you define that?

Hobbie servos are usually rated by the amount of time it takes to travel 60 degrees. Most take around 0.20 seconds to travel 60 degrees. You can get faster servos, but the fastest I've seen is about 0.10 second for 60 degrees. The other servo rating is torque, usually measured in inch/ounces.

The Basic Stamp needs only a millisecond or two to do most functions. It's the fastest part in the equation.

The slowest part of the process would be getting the reading from the gyro or acclerameter. The easiest is serial communication. It's also the slowest.

So it comes down to: how much mass do you need to move, how far do you have to move it and at what speed?

If you add the time it takes for a gyro reading, processor processing, and servo reacting you could spend 0.10 to 0.20 seconds. Is that acceptable?

Daryl

AMG AV
10-19-2008, 04:41 AM
Sorry for the long delay in posting. We have been very busy and not enough time for this project, hence the reason we need outside help.

The .10 to .20 response times are very reasonable, but I have no way of knowing if that is fast enough. I would imagine that is sufficient since the movement we are trying to counter is not faster that those times. Moving platforms almost always have some type of gas or electric motor that produces vibration and that would be measured in Hz, right? Since servo movement is in sec/60d that is somewhat of a Hz measurement? The reason I ask is purely curiosity, not necessarily related to the servo control. Anyway, we are trying to counteract unwanted movements in the fastest rates available in this smaller scale.

Anyway, the hobby servos are fast enough but I would like to have control over the system using a joystick and micro processor that is not part of a RF package like RC tx/rx systems. This whole idea is brought about by the need to remove the remote control process and go back to a wired system that runs on either 6v or 12v. 12v systems are readily available on most of our moving platforms like cars and boats. I suppose we could also disassemble an RC radio TX and take out the joysticks, processor, and hard wire a power source and we have a system ready that uses RC setup parameters used in RC helicopter and airplane stabilization systems, but these do not allow as many options as those in the servo controller software seen in the Parallax modules.

The servo control needs to be able to control 3 axis movement in the pitch, roll, and azimuth and move anywhere from 4 to 8 pounds of equipment. Any ideas? Thanks.