2
\$\begingroup\$

I'm trying to find a good way for measuring the speed/acceleration of a swinging pendulum. I know one way is to use an optical encoder for capturing the rotation positions, but unfortunately there's no room for an optical disc.

I was considering an accelerometer/gyro. Would it be possible? Could I somehow compute the change in pitch? Like with the MPU6050, it can spit out the pitch using its DMP.

Would measuring the change in pitch with an accelerometer/gyro be a good idea?

\$\endgroup\$
  • \$\begingroup\$ Any mass you add to the pendulum will affect its period. Also, if you make a wired connection to the pendulum, the wire will add a small amount of mechanical drag. Whether these things are problems depends on what you are doing. An accelerometer will certainly measure acceleration. Instead of using pitch, just use the acceleration data, since that is what you want. The pitch calculation will be subject to error due to the motion of the reference frame, but the acceleration data will be accurate. The gyro data may be accurate also, but gyros are subject to drift. Might need more processing. \$\endgroup\$ – mkeith Jun 5 '16 at 17:15
0
\$\begingroup\$

The accelerometer will give you exact angle in function as inclinometer, but very slow. The gyro will output a velocity, integrating you get angle, but with offset. A fusion algorithm will give you attitude, accelerometer is correcting the bias of the gyro.

\$\endgroup\$
  • \$\begingroup\$ Aren't accelerometers meant to be very sensitive and responsive? I'm trying a MPU6050 on an Arduino, but I can't seem to notice any lag. Or am I not looking close enough? \$\endgroup\$ – John M. Jun 4 '16 at 16:44
  • \$\begingroup\$ The inclinometer data will not be correct in a motional reference frame. Might still be usable somehow. Will certainly reveal the period of the pendulum, but if the OP wants acceleration, why not just use the acceleration data directly? \$\endgroup\$ – mkeith Jun 5 '16 at 17:18
  • 1
    \$\begingroup\$ @JohnMunroe Perhaps I didn't express in right way. A fusion algorithm acc+gyro will give you the estimated angle even without mounting the device on the axis of movement, even if you mount the device on the pendulum mass. Or you can put the sesnsor on the axis, regardless of exact centering the MEMS with physical axis. \$\endgroup\$ – Marko Buršič Jun 5 '16 at 17:42
1
\$\begingroup\$

Several MEMS accelerometer companies sell eval kits with an accel on a small board with a wireless link and a separate wireless USB receiver. Freescale and Analog Devices are two of them. You can measure acceleration directly with the three-axis accel, and you can check it by using the period if you know the mass distribution of your pendulum.

\$\endgroup\$
  • \$\begingroup\$ Unfortunately I don't know the mass distribution of the pendulum. I'm just trying to plot the velocity or acceleration of the pendulum's bob. \$\endgroup\$ – John M. Jun 4 '16 at 16:05

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.