Inclinometers for self-balancing projects?
modemman
Posts: 41
Hi,
I found these MEMS inclinometer chips recently (on digikey.com):
www.vti.fi/midcom-serveattachmentguid-53b766f2850a127b88947edbf1de536f/SCA61T_inclinometer_datasheet_8261900A.pdf
www.vti.fi/midcom-serveattachmentguid-7f8d2a49c0bdf4fdc95fa02add2ee672/SCA100T_inclinometer_datasheet_8261800A.pdf
These are marketed as "inclinometers" and not accelerometers and the data sheet says:
"The VTI inclinometers are insensitive to vibration, due to their over damped sensing elements, and can withstand mechanical shocks of up to 20000 g."
I'm not microelectronics-savvy enough to tell from the data sheet if they give a fast enough response for a self-balancing application and if the acceleration component will still skew the readings, but I was wandering if these can be used to eliminate the need to combine an accelerometer and a gyro and use complex formulas to create a self-balancing bot (Kalman filters and such). All we really care about is the amount of tilt and how fast the object tilts, no? Isn't that why we use gyro+accelerometer - gyro fast but drifts and doesn't know up/down (as it only senses spin), accelerometer precise and knows down because of the 1g gravity, but reads vibrations as false tilt (as it does what it was made for - sensing accelerations)?
modemman
I found these MEMS inclinometer chips recently (on digikey.com):
www.vti.fi/midcom-serveattachmentguid-53b766f2850a127b88947edbf1de536f/SCA61T_inclinometer_datasheet_8261900A.pdf
www.vti.fi/midcom-serveattachmentguid-7f8d2a49c0bdf4fdc95fa02add2ee672/SCA100T_inclinometer_datasheet_8261800A.pdf
These are marketed as "inclinometers" and not accelerometers and the data sheet says:
"The VTI inclinometers are insensitive to vibration, due to their over damped sensing elements, and can withstand mechanical shocks of up to 20000 g."
I'm not microelectronics-savvy enough to tell from the data sheet if they give a fast enough response for a self-balancing application and if the acceleration component will still skew the readings, but I was wandering if these can be used to eliminate the need to combine an accelerometer and a gyro and use complex formulas to create a self-balancing bot (Kalman filters and such). All we really care about is the amount of tilt and how fast the object tilts, no? Isn't that why we use gyro+accelerometer - gyro fast but drifts and doesn't know up/down (as it only senses spin), accelerometer precise and knows down because of the 1g gravity, but reads vibrations as false tilt (as it does what it was made for - sensing accelerations)?
modemman
Comments
I read the sensor datasheet and wonder if it would suffice for your application. I have tried using the Parallax accelerometer sensor to measure tilt and find that under external acceleration influences [noparse][[/noparse]at a much lower frequency than 'vibration'] it is unable to accurately measure tilt. If the sensor you are considering consists of essentially the same elements then you should ask the manufacturer if it will indeed work for your specific application. The literature does not seem to address this concern.
cheers, David