[AR] FOGgy idea \ ... vanguarduino?
- From: David Weinshenker <daze39@xxxxxxxxxxxxx>
- To: arocket@xxxxxxxxxxxxx
- Date: Mon, 18 Jan 2016 20:46:12 -0800
On 01/17/2016 08:55 PM, Henry Spencer wrote:
For example, maybe there's
some cross-coupling between gyro axes that you can take out by combining
all three gyro channels together into a vector and multiplying by a
correction matrix to give three synthetic channel values.
Thinking about this - AIUI, the hard part of making a "good" FOG isn't
the optical system itself, but subtleties in the winding geometry of
the fiber, required to ensure that the gyro responds only to rotation
about one axis and ignores all rotation about other axes.
If one had three "not quite perfect" (but perhaps cheaper to build) FOG
units, each of which only roughly approximated this condition, could one
generate a "good" 3-axis output from a set of them, processed through
this sort of correction matrix algorithm?
-dave w
Other related posts:
