LISY300 Gyro Output

ajward

Hi All...

I've been experimenting with the above mentioned gyro module using Beau's Gyro DEMOv1_A.spin code from this thread (http://forums.parallax.com/showthread.php/127868-Gyro-lisy300). The gyro is wired correctly and is outputting data to the debug window, but I'm just not sure what it is I'm seeing. When I first ran the program, with the gyro stationary, the display started with "Deg: 0" and increased by 1 every six seconds or so. Rotating the gyro about its Z-axis give predictable results... clockwise the values decrease and counterclockwise the values increase. Rotating about the X or Y axes shows no effect.

I realize the module will some bias, but I thought the calibration subroutine would factor that out, Also, I realize the "stationary" gyro will have some movement due to the earth's rotation, but the value I'm getting seems awfully high for that. As long as the module is not moving on my workbench, its orientation doesn't have any effect on the output.

If anyone has put together this setup, I'd love to hear about your experience. Thank you.

Amanda

Beau Schwabe

The Calibration subroutine is just to get an initial starting value. 1 Deg every 6 seconds seems a little high. The X and Y axis only contribute to the amount of error in the Z axis. Remember, ANY error is going to be accumulative.

ajward

Thanks Beau! Your comment gives me a couple of ideas. Still some output oddities to puzzle out tho'.

My idea is to detect or, possibly, measure the Earth's rotation. Don't know if the Lisy is up to the task, but finding out is part of the fun of learning.

Amanda

Beau Schwabe

Even though the LISY300 Gyro is a single axis Gyro, the orientation of the other axis can still make a difference and often what is interpreted as "noise" contains data. Often people will truncate the noise or seemingly random fluctuation. Yes, some of it is random, but true random noise will bias to 50% on and 50% off. If there is data in there it will be influenced to go more one direction or the other. Also, any time that you can't read the data, i.e. processing data sample, the missed data will miss it's chance to influence the bias. To detect the Earths rotation at 15 deg per hour you need to be exactly in line with the Earths rotational axis, if not the results will be slightly off. Think of an analogy of paddling in a boat. If both ores move at the same rate, spend the same time in the water, and are at the same angle, then you move in a straight line. If one ore is slightly different from the other in any way, you have a slight migration to the left or to the right. The same could be said if the alignment to the axis that you are wanting to measure isn't exactly the same. In the real world this is virtually impossible to do, so there is going to be some sort of unwanted drift. If the drift is constant then you can correct for it over time. If it is not then combining the efforts of multiple sensors comes into play where each sensor contributes to the overall goal. The sensors can be other gyros, accelerometers, or even a compass.... others include a horizon detector.... each of which can lock in a certain way to the environment and hopefully fill in the "gaps" where other sensors weren't able to.

Just keep in mind that reading these kind of sensors can be tricky and you can often find yourself in a circular reference puzzle.







Reference:
http://www.mouser.com/ds/2/321/27922-LISY300GyroscopeModuleV1.0-23580.pdf