SparkFun Forums

[SFE-Robert]

Well, most people that have more specific requirements (such as 6 degrees of freedom) would just get to breakout boards and read them separately.

[esklar81]

It just doesn't have what I need....It's a fine board and I'm sure the UAV people love it to bits. Even they must realize they need a magnometer along with their gyros so that they can combine it with a GPS breakout for navigation.

What part of "and HMC5843 (triple-axis magnetometer)" fails to indicate that it has a magnetometer? Is the problem, perhaps, that you've been searching for "magnometer" instead of "magnetometer"?

BTW, I'm curious as to why you believe you need to know compass heading, but not linear acceleration, to stabilize an image.

I'm naively surprised more people haven't requested this feature.

I think "naively" may be a bit of an understatement.
Eric

[sebmadgwick]

VERY cool. Great job. Any more details about this?

Thank you. Details (report, source code ect.) on the sensor fusion algorithm are available here: http://code.google.com/p/imumargalgorithm30042010sohm/
The camera control demo is as simple as: convert quaternion into pitch/roll/yaw angles then set camera tilt/roll/pan servo angles (respectively) to match. For stabilisation, I set servo angles to the negative values.

...I need that more than I need accelerometers.

I am assuming you wish to stabilise angular position. You will need both accelerometers and gyroscopes need both to obtain an absolute measurement of angular position (attitude). That is unless you have an alternative reference measurement of angular position (?).

Even they must realize they need a magnometer along with their gyros so that they can combine it with a GPS breakout for navigation.
I'm naively surprised more people haven't requested this feature.

I've been rattling my brains trying to think of an application which would only demand gyro and magnetometer... The benefit of fusing magnetometer data with gyro data is to compensate for brief periods of soft-iron distortion and to increase the dynamic accuracy of measurements of the earth’s flux. But then, any practical application would probably require ‘tilt compensation’ and so incorporate an accelerometer.

Am I missing the point :s